Building and Environment 121 (2017) 119e129

Contents lists available at ScienceDirect

Building and Environment

journal homepage: www.elsevier.com/locate/buildenv

A review of simulation-based urban form generation and optimization

for energy-driven urban design
Zhongming Shi a, b, *, Jimeno A. Fonseca a, b, Arno Schlueter a, b
a
Future Cities Laboratory, Singapore-ETH Centre, 1 Create Way, CREATE Tower, Singapore 138602, Singapore
b
Architecture and Building Systems, ETH Zurich, Stefano-Franscini-Platz 1, Zurich 8093, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: This paper first defines the concept of energy-driven urban design. It aims to reveal synergies and trade-
Received 5 February 2017 offs that may arise while designing urban areas for better energy performance. To facilitate urban
Received in revised form planners and designers tackle these problems at the early stage of their work, this paper proposes the
26 April 2017
idea of simulation-based urban form generation and optimization modeling. It connects parametric
Accepted 8 May 2017
Available online 9 May 2017
models of urban form generation to an optimization engine coupled with a widely available program of
energy systems.
To build up the model of simulation-based urban form generation and optimization modeling, this
Keywords:
Urban form generation modeling
paper reviews the state-of-the-art of simulation-based design generation and optimization modeling and
Simulation-based optimization modeling discusses its application on energy-driven urban design at the district scale. The paper compares the
Generative methods main generative methods and presents their limitations and advantages to aid energy-driven urban
Urban morphology design. For the urban form generation modeling, the paper also reviews the most relevant approaches to
Energy-driven urban design urban morphology. These approaches help to define the urban elements for the urban form generation.
Most of the existing design generation and optimization models are observed to consist of a workflow,
a generative method, and a series of generation constraints. Based on this, the paper proposes a model of
simulation-based urban form generation and optimization modeling for energy-driven urban design. The
model consists of a workflow with three steps, a collection step, the generation step, and the optimi-
zation step. The constraints yet need to be defined. At the district scale, the model also has to work at an
appropriate resolution and precision.
© 2017 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2. Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.1. Urban morphology and urban form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.2. Urban energy system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.3. Layer of energy systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.4. Energy-driven urban design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.5. Simulation-based urban form generation and optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3. A review of design generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3.1. Workflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.2. Generative methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4. A review of approaches to urban morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.1. Conzenian (Historico-geographical approach) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.2. City as flows (Complex-systematic approach) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4.3. Typology (Typological approach) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

* Corresponding author. Future Cities Laboratory, Singapore-ETH Centre, #06-01,

1 Create Way, CREATE Tower, Singapore 138602, Singapore.
E-mail address: shi@arch.ethz.ch (Z. Shi).

http://dx.doi.org/10.1016/j.buildenv.2017.05.006
0360-1323/© 2017 Elsevier Ltd. All rights reserved.
120 Z. Shi et al. / Building and Environment 121 (2017) 119e129

4.4. Functional zoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

4.5. Defining the constraints that build up an urban design prototype . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.5.1. Generic urban design constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.5.2. Vernacular urban design constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5. A review of urban form optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.1. Algorithms for urban form optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.2. Tools for urban energy simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.3. Optimization objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6. A new model for energy-driven urban design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7. Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

1. Introduction driven urban design in Section 2, we review the existing models

of design generation modeling, their workflows, and the generative
54% of the world's population lived in urban areas in 2014, and design methods in Section 3, relevant approaches to urban
the number is expected to hit 66% by 2050 [1]. With the current morphology in Section 4 and approaches to urban form optimiza-
prevailing population growth and urbanization rate, by 2030, the tion in Section 5. As a synthesis of the reviews, Section 6 proposes a
global urban area will triple, compared to that of the beginning of new approach towards energy-driven urban design. Section 7
the Twentieth Century [2]. Cities consume around three-quarters of contains the summary and future research outlook.
global primary energy and account for nearly 60% of the world's
greenhouse gas emissions [3]. How the urban areas are built has a 2. Definitions
substantial impact on the urban energy performance of the present
and the future [4]. In recent years, urban energy studies started to In the following section, we define relevant aspects and termi-
embrace the tendencies to focus on the intermediate scale between nology to set the stage for the proposal of energy-driven urban
the city scale and the building scale [5], namely the district scale. design.
Urban design and sometimes urban planning work on this scale, in
the process of urban development. In other words, energy consid- 2.1. Urban morphology and urban form
erations are better and effective to be considered at the beginning
of an urban development process [6]. Urban morphology connects the technological aspect with the
Recent attempts to incorporate energy considerations into ur- formal and various other aspects of cities [12]. Marshall and Çal-
ban planning and design have been made. Fonseca et al. [7] created işkan [13] summarized the definitions of urban morphology in
four different plausible urban transformation design scenarios of a different research contexts. This paper adopts the most basic defi-
small site in Zug, Switzerland, to find out the best zoning strategy of nitions by Cowan [14], Lozano [15] and the International Seminar
the four. Though the design team was comprised of experts like on Urban Morphology [16] e urban morphology is “the study of the
architects, urban designers, engineers, the four scenarios have been physical urban form” and the various factors (e.g. the people and
intuitively produced. However, new patterns of development and the process) that govern and influence form”.
technology render the traditional intuitive methods of designers In this paper, by urban form, besides physical geometries, we
not adequate to manage the interdependencies between various also refer to the urban program and open spaces. The links between
aspects of urban form and energy systems [8]. Optimization urban form and building energy performance have been estab-
modeling is a promising method of minimum energy bench- lished at the district scale [17]e[26]. These links mainly affect the
marking for a new urban development [9]. A computationally energy performance through manipulating the physical geometries
optimized design solution can be very different from and much for lower operational energy demand. However, there is a lack of
more energy efficient than those based on the designer's subjective the knowledge of the interdependencies between urban form and
intuition [10]. energy systems.
Simulation-based optimization modeling is an efficient way to Some energy technologies have enabled new possibilities of the
meet several stringent requirements of high-performance building urban form. For example, a district cooling system can free the
design [11]. Combining a simulation program with an algorithmic building rooftops for more architectural design options, like an
optimization engine has become popular since the late 2000s [11]. infinity pool or a sky garden. Vice versa, to maximize the energy
The results of building simulation programs serve as the optimi- performance, some energy technologies enforce some specific re-
zation's objective function. Likewise, beyond a single building, quirements (e.g. density, building program) on the urban form.
simulation-based optimization can be applied to the district scale Moreover, building rooftops and façades can be equipped with solar
when coupling with an urban energy simulation program. Besides, panels for in-city energy generation. The layout and heights of
since nothing has been formed yet at the beginning stage of urban buildings may affect the mutual shading and further influence the
planning, this urban energy simulation program and the optimi- solar-harvesting potential. Some studies have discussed such in-
zation engine require to be connected with a parametric model of terdependencies between building programs [27], density [28] and
urban form generation. urban energy supply infrastructures.
This paper reviews the state-of-the-art of the research and
knowledge on design generation modeling, urban morphology, and 2.2. Urban energy system
optimization methods for urban form optimization modeling.
Consequently, the review is structured in three parts. After defining The definition of urban energy system has constantly been
the terminologies relevant for setting up the concept of energy- evolving. In Urban Energy Systems, based on Jaccard's 2005
 Z. Shi et al. / Building and Environment 121 (2017) 119e129
definition [29], Keirstead et al. defined urban energy system as “the
combined processes of acquiring and using energy to satisfy the
energy service demands of a given urban area” [30]. Bale et al.
consider that an urban energy system is made up of three com-
ponents, e.g. “agents, interacting through networks”, “objects, such
as technologies and infrastructures” and “the environment, which
provides resources” [31]. Based on these two definitions, the urban
energy system at the district scale is composed of agents and ob-
jects of energy transportation, energy storage, energy conversion
(excluding in-city generation), end-service, and in-city generation,
as well as the environment that provides the resources.
2.3. Layer of energy systems
We define the space for the objects of urban energy systems as
the “layer of energy systems”. In any city, the layer of energy sys-
tems is indispensable. It accommodates the energy objects of en-
ergy transportation, energy storage, energy conversion, end-
service, and in-city generation. Prevailing energy objects and
technologies are associated with one or several of these five energy
functions [32]. Usually, the layer of energy systems is hidden from
the city dwellers. For example, the pipelines and cables are buried
underneath the streets in cities.
2.4. Energy-driven urban design
One of the main tasks of urban designers is “place-making” [33],
[34]. Energy systems influence the quality of the place, like the
thermal comfort. On this basis, we introduce the concept of energy-
driven urban design. Energy-driven urban design aims at providing
urban configurations that decrease energy demand, particularly
considers and responds to interdependencies between the energy
systems and the urban form for higher energy performance, and
makes the place for the layer of energy systems, while maintaining
a high level of human comfort. The energy performance of energy
systems is a general concept that refers to the efficiency and the
cost of the energy system as well as the renewable energy ratio
attained and the emission caused. Fig. 1 illustrates the concept of
energy-driven urban design. Fonseca proposed a similar concept of
performance-based urban design. It addresses the tradeoffs be-
tween urban forms and energy performance together with the
location, costs, and size of the energy infrastructure [35].
The links between urban form and energy demand as well as the
urban energy supply system can be generalized as energy-driven
urban design constraints. We define two categories of such con-
straints, active and passive energy-driven urban design constraints.
Passive energy-driven urban design constraints shape the urban
form to “work on its own” to save energy. For example, to decrease
the heating demand, a passive constraint may manipulate the area
of façades or the orientation of the building for more solar heat
gain. Active energy-driven urban design constraints shape the ur-
ban form to accommodate the requirements of active energy in-
frastructures or technologies. For example, to increase solar
utilization, an active constraint may maximize the area of façades
or rooftops of adequate solar radiation for PV panels.
2.5. Simulation-based urban form generation and optimization
A model of simulation-based design generation and optimiza-
tion consists of three parts, as illustrated in Fig. 2. The simulation
program is combined with the optimization engine as the fitness
function. The optimization engine optimizes the output of the
parametric model of design generation. The model generates
design based on a sequence of processing steps [36] and generative
design constraints. We define this sequence of processing steps as
121
the workflow of urban form generation modeling.
3. A review of design generation
In this section, we review works of city, urban or building design
generation relevant in the context of energy-driven urban design.
The keywords used to search for this studies include generative
urban design, generative design, and design generation. Among the
results, we selected the most interrelated, relevant, and widely-
cited papers since 2000.
Beira~o's model (or the CItyMaker) generates urban form at the
city scale. Its model, called the City Induction generation model,
links three modules, corresponding to three separate urban design
processes [37]e[39]. These are formulation module (analysis),
generation module and evaluation module. The formulation mod-
ule reads the given urban context of the target site [37], [40], and
the evaluation module evaluates and leads the process of urban
form generation to meet the goal of design [37], [41]. The genera-
tion process of CItyMaker adopts the generative method of shape
grammar [42]. Beira ~o et al. defined Urban Induction Pattern (UIP) as
general repetitive urban design moves (the rules for shape
grammar) that can be applied independently from context [37].
Beia~o et al. defined four sets of UIPs through interviews with the
designers of four case studies [37]. In the generation module, UIPs
mimic the general design moves of these designers. Besides the
pre-set UIPs, CityMaker allows customizable UIPs for personalized
design.
Koenig's model generates urban form at the district scale. The
model comprises four levels (steps), the information level, the site
development level, the building level and the optimization level
[43]. The information level stores and retrieves local information,
which includes geographical as well as social and economic con-
ditions of the site. Both the site development level and the building
level generate forms. The former, generates the road networks
based on Humpert's six basic urban patterns [44], while the latter
extrudes the two-dimensional building footprints into three-
dimensional building structures using cellular automata [45]. The
optimization level optimizes the generated forms using genetic
algorithm [43]. Knecht and Koenig's model [46] generates plots and
buildings with a given street network. The buildings are generated
based on four types, row buildings, courtyard buildings, ribbon
buildings, and free-standing blocks [47]. Koenig et al.’s model [48]
generates street networks at the city scale. It adopts a generative
method adapted from the self-sensitive L-system by Parish and
Müller [49]. This model's optimization algorithm uses the choice
value of the network as the fitness function. A street segment's
choice value is, for example, “the number of shortest paths from all
street segments to all others that pass through that segment [48]”.
Rakha and Reinhart's model generates urban form at the district
scale [50]. There are two steps in its workflow, the generation step
and the optimization step. The former generates street patterns and
building masses. The latter evaluates the walkability and optimizes
the design with a genetic algorithm. Different from the two models
above, beyond a perfectly flat site, this model can work on terrain.
However, it has few massing options for 3D buildings and little care
for the void open space.
Caldas et al.’s model works on the building scale and adopts a
Genetic Algorithm [51]for optimization. It optimizes the selected
design variables of an existing building [52]. Caldas et al. provided
an example of using this model for the sizing and placing optimi-
zation of windows on the four façades of an existing office building
for higher energy and thermal performance [52]. Extensive similar
works have been done on optimization of one or several parame-
ters of an existing design prototype [9], [10], [26], [53]e[65].
Stouffs and Janssen's model [47] generates relevant building
122 Z. Shi et al. / Building and Environment 121 (2017) 119e129

Fig. 1. The concept of energy-driven urban design.

Fig. 2. Simulation-based urban form generation and optimization modeling.

data that can be used for the analysis and assessment of urban
planning at an early stage. It functions with consideration of local
conditions, building typologies, and building codes and regulations.
It adopts generative methods of shape grammar.
Luca's [66] model generates urban form in multiple scales, from
district design up to urban and regional design with Agent-Based
Modeling and cellular automata. Though not explicitly described,
the model consists of two steps, one collecting data and one
generating the form. The latter has multiple tasks connected with a
pre-set dataset, and these tasks are arranged in a spatial/temporal/
scale hierarchy (similar to the Historical-Geographical morpho-
logical approach and the resolution in Section 3.1 and Section 3.3).
It does not contain an optimization process. In addition, it, together
with the other three models working in the district or a bigger
scale, lacks the generation of the urban or building program (e.g.
residential, office, retail).
3.1. Workflows
Among these works, Koenig's model [43], Koenig et al.’s model
[48], Rakha et al.’s [50] model, and Caldas et al.’s model [52] fall in
line with the configuration of simulation-based optimization
modeling. Except that of Caldas et al.’s model, the workflows of the
rest have three steps, the Data Collection Step, the Generation Step,
and the Optimization Step. The Data Collection Step collects all
physical, socio-economic data of the environment, reads the
context of the site and user's customized preferences. The Gener-
ation Step generates the urban form with certain generative
methods and constraint sets. The Optimization Step optimizes the
designs produced in the Generation Step and with a genetic algo-
rithm as the optimization engine. The workflow of Caldas et al.’s
model stands out as it starts from an existing design prototype,
while the other workflows both start the generation from an empty
site with or without a context.
Based on these, we propose a general workflow of two parts
with three steps for simulation-based urban form generation and
optimization modeling, as shown in Fig. 3. The Data collection step
and the Generation step make up Part 1: Design Prototype Gener-
ation. The Optimization Step makes up Part 2: Design Optimization.
Part 1 collects customized data for this specific design, like the floor
area ratio. It also contains pre-set default databases, like the
building typologies and constraint sets for design prototype gen-
erations. Part 2 optimizes the urban design prototype generated
through Part 1. The outcomes of this workflow are a series of
optimized urban design solutions for the urban designer to eval-
uate, select and further develop. For energy-driven urban design,
the outcomes are urban design solutions with high energy
performance.
 Z. Shi et al. / Building and Environment 121 (2017) 119e129
Data collection step Generation step
Part 1:
Design Prototype Generation
Fig. 3. A general workflow of design generation and optimization modeling.
3.2. Generative methods
Generative methods have been proven to greatly facilitate the
process of design exploration [67]. Singh and Gu reviewed five
commonly used generative methods, shape grammars, L-system,
cellular automata, Swarm Intelligence (usually utilized by agent-
based modeling [68]), and Genetic Algorithm [69]. All these
generative methods can be found in the works previously intro-
duced. Shape grammar is “a set of shape transformation rules that
are applied recursively to generate a set of designs [42]”. L-system
is a parallel string rewriting - the symbolic representation of the
formal design - mechanism based on a set of production rules [49],
[69]. Cellular automata is “a collection of cells on a grid of specified
shape that evolves over time according to a set of rules driven by
the state of the neighboring cells [45], [70],” [69]. Agent-based
modeling originated from cellular automata [71]. Agent-based
modeling is “a class of computational models for simulating the
actions and interactions of autonomous agents with each other and
the environment” [68]. Genetic Algorithm is “a procedure loosely
based on Darwinian notions of survival of the fittest, where selec-
tion and recombination operators are used among candidate so-
lutions to look for high-performance ones [52], [72],”.
Based on the work of Singh and Gu [69], Table 1 reviews and
compares these five generative methods by components, main
advantages and limitations, design purposes and outcomes,
outcome characteristics and the degree of user intervention.
Different methods, with different advantages and limitations, are
suitable for different design purposes and desired outcomes. In the
generative urban design workflow, the non-optimization methods
work in the Generation step to generate the design prototype, while
the optimization methods work in the Optimization step to opti-
mize the design for better energy performance.
Concerning urban design, shape grammars or L-system can be
used to generate building forms and layouts, road and networks
and plot subdivisions in a hierarchical top-down process. On the
opposite, cellular automata and agent-based modeling can be used
to generate context-sensitive urban massing in a bottom-up pro-
cess. Compared to top-down generative methods, bottom-up ones
have high possibilities of design emergency. Singh and Gu [69]
proposed an integrated framework tested through sample design
scenarios. Coupled generative methods can make up for each
other's drawbacks. For example, bottom-up generative methods,
like agent-based modeling can compensate the drawbacks of top-
down generative methods, like shape grammars. However, the
coupling of different generative methods is sophisticated and
challenging. In some of the models described earlier this section,
coupled methods are adopted as well for different design purposes
[43], [50], [66].
4. A review of approaches to urban morphology
The tools of urban form generation shall produce design options
that are architecturally, geographically, historically and technically
plausible. The Optimization step requires a plausible initial state or
123
Optimization step
Part 2:
Design Optimization
the design prototype from the Generation step to start with [77]. To
generate a plausible initial state or design prototype, an under-
standing of the mechanism of how the city was formed and grew is
a prerequisite. Urban morphology is essential [78] and helps to
understand this mechanism at various scales [79]. From the clas-
sical urban morphological studies to that of now, a series of ap-
proaches prevail in the field. The following section focuses on the
morphological approaches that are relevant to understand the
mechanism of the formation of cities, which constraint the urban
form generation. Each of these approaches provides a different
perspective of understanding how a city is formed. The selection
and categorization of these urban morphological approaches was
based on a recent review work on urban morphology [5]. Also,
additional viewpoints are reviewed from the Journal Urban
Morphology (from 1997 to present) accessed via Journal Online
website of Urban Morphology. Urban Morphology is a journal
curated by the Internal Seminar on Urban Form (ISUF), the inter-
national organization of urban form for researchers and practi-
tioners [80].
4.1. Conzenian (Historico-geographical approach)
Conzenian refers to “an adherent of the doctrines of M.R.G.
Conzen [16] in urban morphology. In this paper, Conzenian is
extended to represent those that belong to the Historico-
geographical approach, in which, time and hierarchy play the key
role. With a descriptive purpose and an aim of developing a theory
of city building, Historico-geographical approach concerns how
cities are built and why [81], [82]. Usually, cities are dissected into
components (e.g. streets, plots, and buildings). All the components
come in a hierarchical-chronological order. The hierarchy and the
lifespan of these components define that component's impact in
the urban context.
Conzen's tripartite division of urban landscape includes the
ground plan (streets, plots, and block plans of buildings), the
building fabric and land, and building utilization [83], [84]. Simi-
larly, Scheer [85] layered urban form components into five layers
based on their expected rates of change. From the most enduring to
the most ephemeral, they are the site, superstructure (e.g. highway,
boundaries existed before urban settlement), infill (e.g. finer paths,
plots), buildings, and objects (e.g. vegetation, fences). In a recent
paper, Scheer [86] reiterated the three basic urban elements as built
form (including objects, buildings, and infrastructure), boundary
matrix (including plots and paths), and land. Beira ~o et al. defined
City Ontology [87]. Ontology is a concept borrowed from the field of
computer science to describe the definitions and interrelationships
of entities that are truly fundamental to a particular domain. Net-
works, blocks, zones, landscapes and focal points are the five main
classes of the City Ontology [38].
The common elements shared by the three are site (land,
landscapes), paths (some superstructures and infill, and networks),
plots (infill), buildings (block plan of buildings and the building
fabric), and objects. Both building utilization and zones will be
addressed in detail later in “building program”. Though urban
124 Z. Shi et al. / Building and Environment 121 (2017) 119e129

Table 1
Comparison of five generative design methods for urban design purposes, modified from Singh et al.'s work [69].

Generative Shape Grammars L-system Cellular Automata Agent-based Modeling Genetic Algorithm
design methods

Components Initial shape, shape rules,
terminal shapes,
operators
Rule application Usually, one rule each
time
Main advantages Visually defined, usually
top-down
Main limitations Incrementally local
behaviors are lost
Design purpose Space layout, patterns
Design outcome Architectural styles,
roads and networks, plot
subdivision
Outcome Repetitive, geometrical,
characteristics validation required
User intervention High, evaluation of
outcomes
References [38], [42]
Initial symbols, rules Grid, Cells, Rules Agents, rules, environment Genes, Chromosomes, Population,
Fitness function
Many rules applied at Parallel computational process Usually, one rule each time Usually one operation each time
the same time
Symbolic strings, usually Bottom-up, context-sensitive Bottom-up, social behavior, Optimization, improvement
top-down simultaneous
Incrementally local Restricted to the geometry of The outcome is difficult to Usually computational expensive
behaviors are lost the cells predict and assess and mono-objective
Patterns Grid-based Way/path-finding Optimization
Architectural styles, Urban massing Traffic flow, walkways, public Optimized solutions
roads and networks, plot space, sometimes for
subdivision geometric design
Repetitive, organic, Geometrical, emergent, Emergent, normative Optimized, usually satisfactory
validation required normative, context-complied, solutions
usually satisfactory
High, evaluation of Low, once cell dimension, state Varies with applications Low, once genotypes, fitness
outcomes rules and initial rules are functions and termination
defined conditions are defined
[49], [73] [45], [74] [71], [75] [51], [76]

blocks are important to the city [88], since they automatically 4.3. Typology (Typological approach)
defined as streets are laid out, here, it is not considered as a
fundamental urban component. Conzenian approaches provide an Both Process typological approach and Rossi's typological
inevitable useful way of understanding how the cities were built approach are two typological morphological approaches. The
and organized [75], [89]. notion of “typology” in these two approaches is similar and related,
but not identical. With a prescriptive purpose and an aim of
developing a theory of city design, Process typological approach
focuses on how cities should be built [81], [82]. With typological
study as the main method, Rossi's typological approach highlights
the importance of space in the development of cities.
4.2. City as flows (Complex-systematic approach)
In the Architecture of the City, Rossi defined typology as elements
that cannot be further reduced [93]. Rossi's typology, not a physical
“City as flows” comes from Michael Batty's editorial “Cities as
form itself, reflects people's way of living. Rossi advocated “a
flows, cities of flows” [90]. In a city, there are flows of human and
remembering of the past” and a city should always be built with
flows of resources. The flows of resources include both natural (e.g.
that “typology of the city”. This is true in many European city cores
sunlight, wind, water) and human-related flows (goods, informa-
today. The ever changing “people's way of living” has successfully
tion, energy, and waste) [91]. The “flows” refer to the changes
fit into the static “typology of the city”. In other words, while the
taking place between the fixed points defined by location and time
physical form of the city remains, the program behind that form has
in both Eulerian and Lagrangian frames of reference [90]. Batty gave
constantly been evolving. The “procedural typology” [94] focuses
a simple distinction between these two frames of reference: the
on building types as the element root of urban form [81]. In the
former is “a bit like watching a place change as people come and
Process typological approach, the typology is somewhat the
go”, while the latter is “more like watching a person traveling
materialized version of Rossi's. Caniggia, the establisher of the
through time from place to place” [90]. Influenced by the science of
method of process typology, connected the phases of a city's history
complexity and theory of systems, Complex-systematic approach
with basic typological processes [5], [95].
concerns the interrelationships between components and these
For urban designers, typology is “an admirable tool” [96]. In
individual components working as a whole. Different from other
laying out the future of cities in Europe, from the late 1960se1970s,
morphological approaches, the Complex-systematic approach
Krier encouraged to use the typological study to guide designs,
looks at a city as a network of flows rather than locations [75], [92].
reintegrate the fragmented parts of the city and imitate the best of
In addition, it works as an approach to synthesis rather than
pre-industrial cities [88]. However, as people's way of living is
analysis. Synthesis addresses the relations in between individual
changing much more dramatically than it was in the past, the ty-
components and them as a whole while analysis tends to describe
pology of the city will no longer host that for eternity and the el-
these individual components.
ements of urban form is under continuous transformation and
“Flows” work as a “hidden hand” and shape the “space of place”,
replacement [81]. Thus, new typologies will emerge, rather than
the “space of flow”, and the “space of systems”. Therefore, at the
being wholly invented [97], reflecting the new people's way of
beginning of a project, the designer needs to figure out the flows
living. The longevity and emergence of typologies depend on
that matter and the agents along the flow, both locally and globally.
multiple factors that can alter people's way of living, like the new
Opposite to the Conzenian hierarchical “top-down” approaches,
technologies [98]. To generate future urban forms, one should focus
“cities as flows” provides a bottom-up perspective of understand-
on the prediction of the future typologies by tracing the evolution
ing how cities are built or grow and why. In the process of urban
of people's way of living and the reasons behind.
form generation, incorporating the bottom-up approach of “cities
as flows” may compensate the drawbacks of the strict hierarchical
and top-down Conzenian approaches.
 Z. Shi et al. / Building and Environment 121 (2017) 119e129
4.4. Functional zoning
Building program usually lays out the activities and schedules
within a building. After decades of zoning-practicing and zoning-
criticizing [88], [99], mixed-use has become a popular strategy for
programs in the district scale. For example, Krier pointed out that
zoning can be subsidized by a community of residence, work, and
leisure all within walking distance and no public transport can
effectively reduce the material and energy waste caused by zoning
[88].
The Urban Land Institute (ULI) defined that a mixed-use project
should have at least three significant revenue-producing and
mutually supporting uses with physically and functionally inte-
grated project components in a coherent plan [100]. Compared to
the traditional zoning code, a mixed-use project raises questions
like how “mixed” should the project be? Where to put the apart-
ments, the retails, and the offices?
Since the urban environment is dynamic and ever changing, the
fittest program of a certain place should adapt as well. Since it is not
rare to see buildings retrofitted and re-programmed [101], building
program is considered more ephemeral than the “buildings”. Thus,
the form and the program of a building are not necessarily related:
based on the function of a building, its form cannot always be
presumed [85] and vice versa. This is also in line with the idea
Rossi's typology. “Less and less surface has to represent more and
more interior activity”, which Koolhaas defined as “Lobotomy” in
the Delirious New York [102].
4.5. Defining the constraints that build up an urban design
prototype
4.5.1. Generic urban design constraints
Based on the discussion on “Conzenian”, “cities as flows”, and
“building program”, we summarize seven components of an urban
design prototype. From rough and enduring to fine and ephemeral,
they are site, streets, plots, buildings, programs, objects, and flows.
Site refers to natural features like landscape, waters, vegetation,
and climate condition. Paths include transport networks of human,
energy, goods, water and waste, like street, rail, canal, and pipes.
Plots mean block subdivisions. Buildings are endurable built
structures with assigned programs, which also include open spaces
with assigned activities. Objects mean non-occupied constructions,
less endurable structures, and planted vegetation. Vegetation ap-
pears both in site and objects, as it can be affected by both natural
processes and human interventions [103]. For example, trees can be
easily wiped out for constructions, while sometimes, buildings can
be designed with respect to the existing trees [104]. The move-
ments of artificial and natural resources (e.g. electricity, food, and
water), as well as human, constitute the urban flows.
Moreover, the forms of the solid components define void open
spaces. The design of such open public spaces is one of the main
tasks of urban design. Such open public spaces include an un-
buildable water body of the site, a lively street and a plaza enclosed
by the immediate buildings around. We define this urban config-
uration of the seven urban components and the open spaces as the
generic urban design constraints.
Resolution is fundamental to urban morphological research
[81]. In the Conzenian hierarchy, there are four levels of resolutions
of urban form, basic components like the building/lot, the street/
block and larger components like the districts/tissues/cities, the
regions [81], [85]. However, to study the influence of urban form
energy performance at the district scale, the level of resolution has
to go finer. For example, window-to-wall ratio and envelope ma-
terial are at a finer resolution than a building volume. Both of them
can have a substantial influence on the energy performance [105],
125
[106].
4.5.2. Vernacular urban design constraints
Kropf [91] identified more general aspects of urban form
through various morphological approaches. Among them is the
social and economic context or local culture. It is important as it,
together with physical context (site), creates the environment for
the city to generate and grow. Other aspects on interrelations be-
tween human and the urban physical features include social and
economic context, control (ownership), intention/design, con-
struction, and perception. We define these, together with the de-
signers' experiences, the common present and future practices of
design (Typological approach), and the local governmental regu-
latory codes as vernacular urban design constraints.
5. A review of urban form optimization
Recent reviews on the optimization algorithms for energy-
related research and the programs of urban energy simulation are
present. Based on these, this section performs the literature studies
on the algorithms for urban form optimization and programs of
urban energy simulation as follows. For the metrics of urban form,
the keywords used include urban form measurement and urban
form metrics.
5.1. Algorithms for urban form optimization
A series of papers have reviewed and compared different opti-
mization algorithms [11], [64], [107]. Among them are the widely
used evolutionary optimization family (e.g. Genetic Algorithm [51],
[76]), Swarm Intelligence [108], [109], Hybrid family (e.g. CMA-ES/
HDE [110]), etc. According to a recent review of simulation-based
optimization methods for building performance analysis [11]: the
most popular optimization algorithms are GA and PSO (Particle
swarm optimization). However, no single optimization algorithm
can be the global “fittest” [11], [64], or called as “no free lunch
theorem [111]”. For example, Ka €mpf et al. [107] proved that the
CMA-ES/HDE has better performance than the PSO-HI in solving
problems. However, Nguyen et al. [11] pointed out that PSO-HI
performs better when the given conditions change. On a case by
case basis, the user makes the decision on selecting a suitable
optimization algorithm [64]. In the works with optimization in
Section 3, optimization algorithms like hybrid CMA-ES/HDE [53],
[54], and several Genetic Algorithms [56] were adopted.
Some real world design problems can be much more compli-
cated and conflicting [65], [112], such as minimum energy con-
sumption versus maximum thermal comfort, etc [11]. This calls for
the application of multi-objective optimization algorithm (MOOA),
though 60% of the building optimization research adopted single-
objective approach [113]. Program coupling and computational
cost [11], as well as the weight of each object in an MOOA, are also
huge challenges for simulation-based optimization methods.
5.2. Tools for urban energy simulation
To perform urban form optimization for energy-driven urban
design at the district scale, a program (or a tool) capable of both
simulation for energy demand and district energy supply system
simulation is required. Energy demand includes operational and
embodied energy consumption. Allegrini has reviewed 20 tools
available for urban energy system simulation and analysis [114].
According to a comparison of these two tools by their capabilities,
no single tool is able to meet all the requirements. Allegrini pro-
poses an approach of integrating different models that address
particular problems, and the Holistic Urban Energy Simulation
126 Z. Shi et al. / Building and Environment 121 (2017) 119e129

(HUES) Platform [115] is on this direction. In recent years, a toolkit
called, the City Energy Analyst (CEA), capable of detailed district
energy supply system design optimization and energy demand
(operational and embodied) simulation is emerging [116].
5.3. Optimization objectives
Besides urban energy performance like carbon emissions, pri-
mary energy consumption, and costs, other urban design perfor-
mances (e.g. walkability, outdoor thermal comfort) matter as well.
This requires a series of metrics of urban form measurement that
are associated with the evaluation of various urban design attri-
butes. Song et al. proposed methods of measurement in five classes:
street network connectivity, density, land use mix, accessibility, and
pedestrian walkability [117], [118]. More in detail, measurements of
urban forms at the district scale include but are not limited to block
length [119], block density [120], space syntax [121], street density
[122], pedestrian route directness [123], floor area ratio [124],
building block coverage [124], density gradient [125], [126], occu-
pant density [124], [126], depth of lot [127], lot size [127], sky view
factor [17], commercial/retail/park distance [117], commercial/
transit pedestrian accessibility [117], surface to volume ratio [17],
land use mix [128], [129], walk/transit/bike score [130], [131],
Predicted Mean Vote (PMV) [132]. Table 2 reviews these metrics of
urban form measurement by their brief description. One or some of
these metrics can either be the fitness objectives for the optimi-
zation engines or simply design indicators for user's own judgment.
6. A new model for energy-driven urban design
Through this paper, we first present the definition of energy-
driven urban design. In addition, we define the concept of the
layer of energy systems and active energy-driven urban design
constraints to link the energy infrastructures of urban energy sys-
tems with urban form. Then, recent models of design generation
modeling were reviewed. A workflow comprising three steps for
urban form generation and optimization was generalized. The three
steps are the Data collection step, the Generation step, and the
Optimization step. Urban morphological studies were reviewed for
the generic urban design constraints that make up the urban design
prototype in the Generation Step. Furthermore, studies on non-
optimization generative methods for the Generation step, and
optimization algorithms and energy simulation program for the
Optimization step were discussed. Fig. 4 depicts the brief model of
simulation-based urban form generation and optimization
modeling of energy-driven urban design we concluded from the
review in this paper.
The Data collection step collects all the available information
from the user and the context. These data include government
regulations, user requirements, energy infrastructures, weather
information, socio-economic conditions, like the energy cost,
environmental quality, like the indoor temperature set point,
infiltration rate, and preliminary designs, if any. The updates of
some of these data will lead to updates of the relevant vernacular
urban design constraints. When some of the data cannot be ac-
quired from the user or the context, the Data collection step will call
the pre-set databases. The pre-set database contains the default
data required by the Generation step.
The Generation step generates an energy-driven urban design
prototype with the generic urban design constraints, the vernacular
urban design constraints, and both the active and passive energy-
driven urban design constraints. Through a comparison of the
four common non-optimization generative design methods in
Section 3.2, L-system or its adapted version has a higher potential
here for the urban form generation with the four sets of top-down
urban form generation constraints defined. However, the genera-
tion of different urban components may call for the coupling of
different generative methods as each of the generative methods has
its advantages and limitations.
The Optimization step is simulation-based. It couples an energy
simulation program and an algorithmic optimization engine.
Because of the “no free lunch theorem”, it requires research on the
selection of a suitable optimization algorithm based on the specific
conditions here. The main objective of energy-driven urban design
is to improve the energy performance. One potential urban energy
simulation program that is able to simulate energy demand
(operational and embodied) and detailed district energy supply
system is the City Energy Analyst (CEA). Besides energy perfor-
mance, other urban design metrics valued by the urban designers
can also be considered as the fitness functions of MOOA.
In computational urban form generation and optimization
modeling, besides resolution, precision is of great importance.
Precision means, at a certain level of resolution, the quantitative
exactness of the component's parameters. For example, if the

Table 2
Metrics of urban form measurement at the district scale.

Measurement Brief descriptions References

Block Length Length of the longest edge of the block [119]

Block Density The number of blocks divided by the area of the land [120]
Space Syntax-Connectivity The number of immediate neighbors that are directly connected to a space [121]
Street Density Linear length of streets divided by the area of the land [122]
Pedestrian Route Directness Route distance divided by straight-line distance between two points [123]
Floor Area Ratio Gross built floor area on divided by the parcel area [124]
Building Block Coverage Area of the ground floor building footprints divided by the area of the block [124,126]
Density Gradient “the rate at which density falls (according to distance) from the location of reference” [125,126]
Occupant Density Number of occupants in an individual habitable unit divided by the total floor area [124,126]
Depth of Lot Distance from the center of the plot to nearest parcel [127]
Lot Size Width, length, perimeter and area of the lot [127]
Sky View Factor The openness of the urban text to the sky [17]
Com/Tran/Park Distance Median distance to the nearest commercial uses/transit/parks [117]
Com/Tran Pedestrian Accessibility Percentage of dwelling units within walking distance from commercial uses/transit [117]
Surface to Volume Ratio (shape factor) Total surface area of buildings (façade plus roofs) divided by the volume [17]
Land Use Mix It measures the land us mixture [128]
Urban Form Index Evaluation of the urban spatial vitality [133]
Walk/Transit/Bike Score Measurement of how pedestrian/bike-friendly and transit-accessible a place is [130,131]
Space Syntax-Integration “Integration is usually indicative to how many people are likely to be in space (Hillier, 1996)” [121]
Predicted Mean Vote (PMV) An ISO standard thermal scale that ranges from Cold (
3) to Hot (þ3) [132]
 Z. Shi et al. / Building and Environment 121 (2017) 119e129
Fig. 4. A model of simulation based urban form generation and optimization modeling for energy-driven urban design (EDUD ¼ Energy-driven urban design).
window-to-wall ratio of a building façade is set to a certain range
(e.g. 0.4e0.6), when the computer is trying to generate the form, a
precision of 0.01 (e.g. 0.41) can be at much more computational
expensive than that of 0.05 (e.g. 0.45), especially when coupling
with many other variables. Also, too many design variables can
cause excessive computations in the Optimization step. It may
require sensitivity analyses to determine the weight of each design
variable's relevance to the urban energy performance.
The outcome of the Optimization step is a series of optimized
energy-driven urban design scenarios with high energy perfor-
mance for the designers to judge and select. However, unlike
buildings, an urban design project takes years, even decades to
realize. Phasing is a common strategy for massive urban develop-
ment. Throughout the time of the development, conditions may
change dramatically. For example, in the successive phases, the cost
of in-city generation infrastructure like PV panels may become
much affordable than that at the beginning. A much more massive
application of PV panels can be expected than the previous phases.
A single run of the workflow is not enough. Changes of boundary
condition like this call for additional runs of the entire workflow.
Prior to the full completion of the entire project, the workflow
needs to be repeated as the conditions continue to change.
7. Summary and outlook
Through literature review and discussion, this paper builds up
the structure of the three-step model of simulation-based urban
form generation and optimization modeling for energy-driven ur-
ban design at the district scale. For the Generation step, we define
four sets of constraints for the generation of energy-driven urban
127
design prototype. The generic urban design constraints are derived
through a review of prevailing morphological approaches. An urban
design prototype is composed of the site, streets, plots, buildings,
programs, objects, flows and open spaces. The determination of the
resolution and precision of the model requires sensitivity analyses.
Vernacular urban design constraints can be acquired from typo-
logical case studies of the target urban context to apply the model.
Besides, the active and passive energy-driven urban design con-
straints, especially the former, are the core of the model. Unlike the
latter, few studies have touched the active energy-driven urban
design constraints, namely the interdependencies between urban
form and urban energy supply systems. Furthermore, to complete
the Optimization step, the challenges are the selection of a suitable
optimization algorithm and urban energy simulation program. It
has to be able to simulate the energy demand (operational and
embodied) and the urban energy supply system needed. A candi-
date, yet still under development, is the City Energy Analyst (CEA).
References
[1] United Nations, World Urbanization Prospects, 2014 Revision, Department of
Economic and Social Affairs, Population Division, New York, 2014.
[2] K.C. Seto, B. Güneralp, L.R. Hutyra, Global forecasts of urban expansion to
2030 and direct impacts on biodiversity and carbon pools, Proc. Natl. Acad.
Sci. U. S. A. 109 (40) (2012) 16083e16088.
[3] UN-Habitat, Energy, 2012 [Online]. Available: http://unhabitat.org/urban-
themes/energy/. Accessed 08 November 2016.
[4] V. Oliveria, M. Silva, Urban form and energy, Urban Morphol. 17 (2013)
181e182.
[5] V. Oliveira, Urban Morphology, Springer International Publishing, Cham,
2016.
[6] IEA ECBCS Annex51-Subtask B, Case Studies on Energy Planning and
Implementation Strategies for Neighborhoods, Quarters and Municipal
128 Z. Shi et al. / Building and Environment 121 (2017) 119e129
Areas, 2012.
[7] J.A. Fonseca, A. Schlueter, Integrated model for characterization of spatio-
temporal building energy consumption patterns in neighborhoods and city
districts, Appl. Energy 142 (Mar. 2015) 247e265.
[8] M.W. Mehaffy, Generative methods in urban design: a progress assessment,
J. Urban. Int. Res. Placemaking Urban Sustain 1 (1) (Mar. 2008) 57e75.
[9] J. Keirstead, N. Shah, Calculating minimum energy urban layouts with
mathematical programming and Monte Carlo analysis techniques, Comput.
Environ. Urban Syst. 35 (5) (Sep. 2011) 368e377.
[10] J.H. Ka€mpf, D. Robinson, Optimisation of building form for solar energy
utilisation using constrained evolutionary algorithms, Energy Build. 42 (6)
(Jun. 2010) 807e814.
[11] A.-T. Nguyen, S. Reiter, P. Rigo, A review on simulation-based optimization
methods applied to building performance analysis, Appl. Energy 113 (Jan.
2014) 1043e1058.
[12] M. Maretto, Sustainable urbanism: the role of urban morphology, Urban
Morphol. 18 (2014) 163e164.
[13] S. Marshall, O. Çalişkan, A Joint framework for urban morphology and design,
Built Environ. 37 (Dec. 2011) 409e426.
[14] R. Cowan, The dictionary of urbanism. Tisbury, Streetwise Press, 2005.
[15] E.E. Lozano, Community Design and the Culture of Cities: the Crossroad and
the Wall, Cambridge University Press, Cambridge; New York, 1990.
[16] International Seminar on Urban From, ISUF: Glossary, 2016 [Online]. Avail-
able: http://www.urbanform.org/glossary.html. Accessed: 17-Nov-2016.
[17] C. Ratti, D. Raydan, K. Steemers, Building form and environmental perfor-
mance: archetypes, analysis and an arid climate, Energy Build. 35 (1) (Jan.
2003) 49e59.
[18] C. Ratti, N. Baker, K. Steemers, Energy consumption and urban texture, En-
ergy Build 37 (7) (Jul. 2005) 762e776.
[19] S. Salat, Energy and bioclimatic efficiency of urban morphologies: towards a
comparative analysis of Asian and European cities, in: Proceedings of the
Conference on Sustainable Building South East Asia, 2007.
[20] S. Salat, Energy loads, CO2 emissions and building stocks: morphologies,
typologies, energy systems and behaviour, Build. Res. Inf. 37 (5e6) (Nov.
2009) 598e609.
[21] S. Salat, L. Bourdic, C. Nowacki, Energy and the Form of Urban Fabric: the
Example of Paris, 2013 presented at the Central Europe towards Sustainable
Building.
[22] L. Adolphe, A simplified model of urban morphology: application to an
analysis of the environmental performance of cities, Environ. Plan. B Plan.
Des. 28 (2) (2001).
[23] P. Rode, C. Keim, G. Robazza, P. Viejo, J. Schofield, Cities and energy: urban
morphology and residential heat-energy demand, Environ. Plan. B Plan. Des.
41 (1) (Feb. 2014) 138e162.
[24] S. Salat, C. Guesne, Energy and Carbon Efficiency of Urban Morphologies. The
Case of Paris, Urban Morphologies Laboratory, CSTB (French Scientific Centre
for Building Research) and ENSMP (Ecole Nationale Superieure des Mines de
Paris), Paris, 2008.
[25] H. Sanaieian, M. Tenpierik, K. van den Linden, F. Mehdizadeh Seraj,
S.M. Mofidi Shemrani, Review of the impact of urban block form on thermal
performance, solar access and ventilation, Renew. Sustain. Energy Rev. 38
(Oct. 2014) 551e560.
[26] G. Caruso, J.H. Ka €mpf, Building shape optimisation to reduce air-conditioning
needs using constrained evolutionary algorithms, Sol. Energy 118 (Aug.
2015) 186e196.
[27] R.E. Best, F. Flager, M.D. Lepech, Modeling and optimization of building mix
and energy supply technology for urban districts, Appl. Energy 159 (Dec.
2015) 161e177.
chal, Establishing links
[28] S. Cajot, N. Schüler, M. Peter, J. Page, A. Koch, F. Mare
for the planning of sustainable districts, in: Presented at the Sustainable Built
Environment (SBE) Regional Conference, Zurich, 2016.
[29] M.K. Jaccard, Sustainable Fossil Fuels: the Unusual Suspect in the Quest for
Clean and Enduring Energy, Cambridge University Press, Cambridge, 2005.
[30] J. Keirstead, N. Shah, Urban Energy Systems: an Integrated Approach, Rout-
ledge, New York, 2013.
[31] C.S.E. Bale, L. Varga, T.J. Foxon, Energy and complexity: new ways forward,
Appl. Energy 138 (Jan. 2015) 150e159.
[32] MuSES, Future Cities Laboratory MuSES Workshop e 16 Sept 2016, 2016
[Online]. Available, http://www.muses.fcl.ethz.ch/2016/10/13/muses-
workshop-16-sept-2016/. Accessed 23 November 2016.
[33] P. Buchanan, What city? A plea for place in the public realm, Archit. Rev. 184
(1101) (1988) 31e41.
[34] The Center for Design Excellence, “Urban Design.” [Online]. Available: http://
www.urbandesign.org/. [Accessed 15 November 2016].
[35] J.A. Fonseca, Energy efficiency strategies in urban communities: modeling,
analysis and assessment, Doctoral Thesis, ETH-Zürich, Zürich, 2016.
[36] C. Schinko, R. Berndt, E. Eggeling, D. Fellner, A scalable rendering framework
for generative 3D content, in: Proceedings of the 19th International ACM
Conference on 3D Web Technologies, New York, NY, USA, 2014, pp. 81e87.
[37] J.N. Beira~o, J.P. Duarte, R. Stouffs, Creating specific grammars with generic
grammars: towards Flexible urban design, Nexus Netw. J. 13 (1) (Apr. 2011)
73e111.
[38] J. Beir~
ao, CItyMaker. Designing grammars for urban design, ABE Archit. Built
Environ. 2 (5) (Oct. 2012) 1e440.
[39] J.P. Duarte, J.N. Beir~ao, N. Montenegro, J. Gil, City Induction: a model for
formulating, generating, and evaluating urban designs, in: SpringerLink,
Springer, Berlin Heidelberg, 2012, pp. 73e98.
[40] N.C. Montenegro, J.P. Duarte, Towards a computational description of urban
patterns, in: Integrating Methods and Techniques: 26th ECAADe Conference
Proceedings, Antwerpen, Belgium, 2008, pp. 239e248 vol. In Architecture “in
computro.
[41] J. Gil, J.P. Duarte, Towards an urban design evaluation framework, in: Inte-
grating Methods and Techniques: 26th ECAADe Conference Proceedings,
Antwerpen, Belgium, 2008 (vol. Architecture “in computro.”).
[42] G. Stiny, Introduction to shape grammars, Environ. Plann B Plann Des. 7
(1980) 343e351.
[43] R. Koenig, Generating urban structures: a method for urban planning sup-
ported by multi-agent systems and cellular automata, Res. Gate 16 (Jan.
2012).
[44] K. Humpert, Das Pha €nomen der Stadt: Berichte aus Forschung und Lehre,
St€adtebauliches Institut der Universita €t Stuttgart, Stuttgart, 1992.
[45] J. von Neumann, The general and logical theory of automata, in: Cerebral
Mechanisms in Behavior-the Hixon Symposium, 1948, Wiley, 1951, pp.
1e41.
[46] K. Knecht, R. Koenig, Automatische Grundstücksumlegung mithilfe von
Unterteilungsalgorithmen und typenbasierte Generierung von Stadt-
strukturen, Bauhaus-Universita €t Weimar, Weimar, 2012.
[47] R. Stouffs, P. Janssen, A Rule-based Generative Analysis Approach for Urban
Planning, 2017.
[48] R. Koenig, L. Treyer, G. Schmitt, Graphical smalltalk with my optimization
system for urban planning tasks, in: Proceedings of the 31st ECAADe Con-
ference, Delft, the Netherlands, 2013 (vol. Computation and Performance).
[49] Y.I.H. Parish, P. Müller, Procedural modeling of cities, in: Proceedings of the
28th Annual Conference on Computer Graphics and Interactive Techniques,
New York, NY, USA, 2001, pp. 301e308.
[50] T. Rakha, C. Reinhart, Generative urban modeling: a design workflow for
walkability-optimised cities, in: Proceedings of SimBuild2012, Madison,
Wisconsin, USA, 2012.
[51] J.H. Holland, Adaptation in Natural and Artificial Systems, Second Edition,
MIT Press, Cambridge, 1992.
[52] L.G. Caldas, L.K. Norford, A design optimization tool based on a genetic al-
gorithm, Autom. Constr. 11 (2) (Feb. 2002) 173e184.
[53] T. Vermeulen, J.H. K€ ampf, B. Beckers, Urban form optimization for the energy
performance of buildings using Citysim, in: Proceedings of CISBAT 2013
Cleantech for Smart Cities and Buildings, Lausanne, Switzerland vol. II, 2013,
pp. 915e920.
[54] J.H. Ka €mpf, D. Robinson, Optimisation of urban energy demand using an
evolutionary algorithm, in: Proceedings of the Eleventh International IBPSA
Conference, Glasgow, Scotland, 2009, pp. 668e673.
[55] J.H. Ka€mpf, M. Montavon, J. Bunyesc, R. Bolliger, D. Robinson, Optimisation of
buildings' solar irradiation availability, Sol. Energy 84 (4) (Apr. 2010)
596e603.
[56] M.J.N. Oliveira Pana ~o, H.J.P. Gonçalves, P.M.C. Ferra ~o, Optimization of the
urban building efficiency potential for mid-latitude climates using a genetic
algorithm approach, Renew. Energy 33 (5) (May 2008) 887e896.
[57] W. Yu, B. Li, H. Jia, M. Zhang, D. Wang, Application of multi-objective genetic
algorithm to optimize energy efficiency and thermal comfort in building
design, Energy Build. 88 (Feb. 2015) 135e143.
[58] W. Wang, R. Zmeureanu, H. Rivard, Applying multi-objective genetic algo-
rithms in green building design optimization, Build. Environ. 40 (11) (Nov.
2005) 1512e1525.
[59] D. Tuhus-Dubrow, M. Krarti, Genetic-algorithm based approach to optimize
building envelope design for residential buildings, Build. Environ. 45 (7) (Jul.
2010) 1574e1581.
[60] J. Gagne, M. Andersen, A generative facade design method based on
daylighting performance goals, J. Build. Perform. Simul. 5 (3) (May 2012)
141e154.
[61] Y.K. Yi, H. Kim, Agent-based geometry optimization with Genetic Algorithm
(GA) for tall apartment's solar right, Sol. Energy 113 (Mar. 2015) 236e250.
[62] K. Suga, S. Kato, K. Hiyama, Structural analysis of Pareto-optimal solution sets
for multi-objective optimization: an application to outer window design
problems using Multiple Objective Genetic Algorithms, Build. Environ. 45 (5)
(May 2010) 1144e1152.
[63] R. Evins, P. Pointer, R. Vaidyanathan, S. Burgess, A case study exploring
regulated energy use in domestic buildings using design-of-experiments and
multi-objective optimisation, Build. Environ. 54 (Aug. 2012) 126e136.
[64] M. Hamdy, A.-T. Nguyen, J.L.M. Hensen, A performance comparison of multi-
objective optimization algorithms for solving nearly-zero-energy-building
design problems, Energy Build. 121 (Jun. 2016) 57e71.
[65] M. Fesanghary, S. Asadi, Z.W. Geem, Design of low-emission and energy-
efficient residential buildings using a multi-objective optimization algo-
rithm, Build. Environ. 49 (Mar. 2012) 245e250.
[66] C. Luca, Generative platform for urban and regional design, Autom. Constr.
16 (1) (Jan. 2007) 70e77.
[67] R. Stouffs, Y. Rafiq, Generative and evolutionary design exploration, Res. Gate
29 (4) (Nov. 2015) 329e331.
[68] A.J. Park, H.H. Tsang, M. Sun, U. Gl€ asser, An agent-based model and
computational framework for counter-terrorism and public safety based on
swarm intelligence a, Secur. Inf. 1 (1) (Dec. 2012) 23.
[69] V. Singh, N. Gu, Towards an integrated generative design framework, Des.
 Z. Shi et al. / Building and Environment 121 (2017) 119e129
Stud. 33 (2) (Mar. 2012) 185e207.
[70] S. Wolfram, A New Kind of Science, Wolfram Media Inc., Champaign, 2002.
[71] L. Chen, Agent-based modeling in urban and architectural research: a brief
literature review, Front. Archit. Res. 1 (2) (Jun. 2012) 166e177.
[72] D.E. Goldberg, Genetic Algorithms in Search, Optimization and Machine
Learning, first ed., Addison-Wesley Longman Publishing Co., Inc., Boston, MA,
USA, 1989.
[73] A. Lindenmayer, Mathematical models for cellular interactions in develop-
ment I. Filaments with one-sided inputs, J. Theor. Biol. 18 (3) (Mar. 1968)
280e299.
[74] M. Batty, Cities and Complexity: Understanding Cities with Cellular
Automata, Agent-based Models and Fractals, MIT Press, Cambridge, 2005.
[75] M. Batty, The New Sciences of Cities, MIT Press, Cambridge, 2013.
[76] L.G. Caldas, An Evolution-based Generative Design System: Using Adaptation
to Shape Architectural Form, PhD Thesis, Massachusetts Institute of Tech-
nology, Cambridge, 2001.
[77] P. Janssen, A generative evolutionary design method, Digit. Creat. 17 (1) (Jan.
2006) 49e63.
[78] K. Kropf, The Handling Characteristics of Urban Form, vol. 93, 2005, pp.
17e18.
[79] P.J. Larkham, Understanding urban form? Urban Des. 93 (2005) 22e24.
[80] International Seminar on Urban From, ISUF - International Seminar on Urban
Form, 2017 [Online]. Available: http://www.urbanform.org/index.html.
Accessed: 25-Apr-2017.
[81] A.V. Moudon, Urban morphology as an emerging interdisciplinary field,
Urban Morphol. (1) (1997) 3e10.
[82] V. Oliveira, C. Monteiro, J. Partanen, A comparative study of urban form,
Urban Morphol. 19 (1) (2015) 73e92.
[83] M.R.G. Conzen, Alnwick, Northumberland: a Study in Town-plan Analysis.
Institute of British Geographers London: Publication 27, George Philip,
London, 1960.
[84] J.W.R. Whitehand, British urban morphology: the Conzenian tradition, Urban
Morphol. 5 (2) (2001) 103e109.
[85] B.C. Scheer, The anatomy of sprawl, Places A Forum Environ. Des. 14 (2) (Fall,
2001) 26e37.
[86] B.C. Scheer, The epistemology of urban morphology, Urban Morphol. 20 (1)
(2016) 5e17.
[87] J. Beirao, R. Stouffs, J. P. Duarte, and N. Montenegro, “The city as a street
system: a street description for a city ontology,” Sigr. 2009-Proc. 13th Congr.
Iberoam. Soc. Digit. Graph. Pp 132e134 (Sao Paulo Braz).
[88] L. Krier, Critiques' and ‘urban components, in: Houses, Palaces, Cities,
Academy Publications, London, 1984.
[89] C. Alexander, A city is not a tree, Archit. Forum 122 (1) (1965) 58e62.
[90] M. Batty, Cities as flows, cities of flows, Environ. Plann B Plann Des. 38 (2)
(2011) 195e196.
[91] K. Kropf, Aspects of urban form, Urban Morphol. 13 (2) (2009) 105e120.
[92] P. Holubec, Conceptualizing the Urban System as a System of Flows,” in
Understanding Complex Urban Systems, Springer International Publishing,
Switzerland, 2016.
, Marsilio, Padova, 1966.
[93] A. Rossi, L’architettura Della Citta
[94] G. Caniggia, G.L. Maffei, Architectural Composition and Building Typology:
Interpreting Basic Building, Alinea Editrice, Florence, 2001.
[95] G. Cataldi, From Muratori to Caniggia: the origins and development of the
Italian school of design typology, Urban Morphol. 7 (1) (2003) 19e34.
[96] C.P. Graves, A Book Review: the Evolution of Urban Form: Typology for
Planners and Architects by Brenda Case Scheer, American Planning Associ-
ation, Chicago, USA, 2011, 129pp. ISBN 978-1-932364-87-3,” Urban Morphol,
vol. 17, no. 1, p. 76, 2013.
[97] B.C. Scheer, The Evolution of Urban Form: Typology for Planners and Ar-
chitects, American Planning Association, Chicago, 2010.
[98] J. Brotchie, P. Newton, P. Hall, P. Nijkamp, The Future of Urban Form: the
Impact of New Technology, Nichols Pub. Co, New York, 1985.
[99] J. Jacobs, The Death and Life of Great American Cities, Reissue ed., Vintage
Books, New York, 1992.
[100] R. Witherspoon, J.P. Abbett, R.M. Gladstone, Urban Land Institute, Mixed-use
Development: New Ways of Land Use, Urban Land Institute, Washington DC,
1976.
[101] E. Dunham-Jones, J. Williamson, Retrofitting Suburbia, Updated Edition:
Urban Design Solutions for Redesigning Suburbs, Wiley & Sons, Inc, Hobo-
ken, 2011.
[102] R. Koolhaas, Delirious New York: a Retroactive Manifesto for Manhattan,
Oxford University Press, Oxford, 1978.
[103] M.I.W. Hopkins, Vegetation as a component of urban form, Urban Morphol.
17 (1) (2013) 57e59.
[104] Architecture News Plus, #30 of 33 Images for Youth Music School Extension
by Miralles Tagliabue EMBT, Youth Music School Extension, 2016 [Online].
Available: http://www.architecturenewsplus.com/project-images/31560.
129
Accessed: 24-Nov-2016.
[105] F. Goia, Search for the optimal window-to-wall ratio in office buildings in
different European climates and the implications on total energy saving
potential, Sol. Energy 132 (Jul. 2016) 467e492.
[106] B. Nghana, F. Tariku, Phase change material's (PCM) impacts on the energy
performance and thermal comfort of buildings in a mild climate, Build. En-
viron. 99 (Apr. 2016) 221e238.
[107] J.H. Ka€mpf, M. Wetter, D. Robinson, A comparison of global optimization
algorithms with standard benchmark functions and real-world applications
using EnergyPlus, J. Build. Perform. Simul. 3 (2) (Jun. 2010) 103e120.
[108] A. Payman, Swarm Intelligence, Jet Propulsion Laboratory, Pasadena, 2004.
[109] Y. Shi, Recent algorithms and applications in Swarm Intelligence research,
IGI Global, 2012.
[110] J.H. Ka€mpf, D. Robinson, A hybrid CMA-ES and HDE optimisation algorithm
with application to solar energy potential, Appl. Soft Comput. 9 (2) (Mar.
2009) 738e745.
[111] D.H. Wolpert, W.G. Macready, No free lunch theorems for optimization, IEEE
Trans. Evol. Comput. 1 (1) (Apr. 1997) 67e82.
[112] M. Hamdy, A. Hasan, K. Siren, Applying a multi-objective optimization
approach for Design of low-emission cost-effective dwellings, Build. Environ.
46 (1) (Jan. 2011) 109e123.
[113] R. Evins, A review of computational optimisation methods applied to sus-
tainable building design, Renew. Sustain. Energy Rev. 22 (Jun. 2013)
230e245.
[114] J. Allegrini, K. Orehounig, G. Mavromatidis, F. Ruesch, V. Dorer, R. Evins,
A review of modelling approaches and tools for the simulation of district-
scale energy systems, Renew. Sustain. Energy Rev. 52 (Dec. 2015)
1391e1404.
[115] Empa, “HUES Platform - Holistic Urban Energy Simulation Platform.” [On-
line]. Available: https://hues.empa.ch/index.php/HUES_Platform. [Accessed:
10-Jan-2017].
[116] J.A. Fonseca, T.-A. Nguyen, A. Schlueter, F. Marechal, City Energy Analyst
(CEA): integrated framework for analysis and optimization of building en-
ergy systems in neighborhoods and city districts, Energy Build. 113 (Feb.
2016) 202e226.
[117] Y. Song, G.-J. Knaap, Measuring urban form: is portland winning the war on
sprawl? J. Am. Plann. Assoc. 70 (2) (Jun. 2004) 210e225.
[118] Y. Song, Smart growth and urban development pattern: a comparative study,
Int. Reg. Sci. Rev. 28 (2) (Apr. 2005) 239e265.
[119] S. Handy, R.G. Paterson, K. Butler, Planning for Street Connectivity: Getting
from Here to There, American Planning Association, Chicago, 2003.
[120] R. Cervero, K. Kockelman, Travel demand and the 3Ds: density, diversity, and
design, Transp. Res. Part Transp. Environ. 2 (3) (Sep. 1997) 199e219.
[121] K. Al_Sayed, A. Turner, B. Hillier, S. Iida, A. Penn, Space Syntax Methodology,
fourth ed., Bartlett School of Architecture, UCL, London, 2014.
[122] S. Handy, Urban form and pedestrian choices: study of austin neighbor-
hoods, Transp. Res. Rec. J. Transp. Res. Board 1552 (Jan. 1996) 135e144.
[123] P. Hess, A. Moudon, M. Snyder, K. Stanilov, Site design and pedestrian travel,
Transp. Res. Rec. J. Transp. Res. Board 1674 (Jan. 1999) 9e19.
[124] C.T. Boyko, R. Cooper, Clarifying and re-conceptualising density, Prog. Plan.
76 (1) (Jul. 2011) 1e61.
[125] P.A. Longley, V. Mesev, Measurement of density gradients and space-filling
in urban systems, Pap. Reg. Sci. 81 (1) (Jan. 2002) 1e28.
[126] V. Cheng, Understanding Density and High Density,” in Designing High-
density Cities: for Social and Environmental Sustainability, Routledge, New
York, 2015.
[127] Y. Song, G.-J. Knaap, Quantitative classification of neighbourhoods: the
neighbourhoods of new single-family homes in the portland metropolitan
area, J. Urban Des. 12 (1) (Feb. 2007) 1e24.
[128] L.D. Frank, G. Pivo, The impacts of mixed use and density on the utilization of
three modes of travel: the single occupant vehicle, transit, and walking,
Transp. Res. Rec. 1466 (1994).
[129] K. Manaugh, T. Kreider, What is mixed use? Presenting an interaction
method for measuring land use mix, J. Transp. Land Use 6 (1) (Apr. 2013)
63e72.
[130] Walk Score, Walk Score Methodology White Paper, 2011 [Online]. Available,
http://www2.WalkScore.com/pdf/WalkScoreMethodology.pdf.
[131] L.J. Carr, S.I. Dunsiger, B.H. Marcus, Validation of Walk Score for estimating
access to walkable amenities, Br. J. Sports Med. 45 (14) (Nov. 2011)
1144e1148.
[132] International Orgainization for Standardization, ISO 7730:2005-Ergonomics
of the Thermal Environment e Analytical Determination and Interpreta-
tion of Thermal Comfort Using Calculation of the PMV and PPD Indices and
Local Thermal Comfort Criteria, 2005 [Online]. Available: https://www.iso.
org/standard/39155.html. Accessed: 24-Apr-2017.
[133] Y. Ye, Urban Form Index for Quantitative Urban Morphology and Urban
Design Analyses, PhD Thesis, University of Hong Kong, Hong Kong, 2015.