sustainability
Article
Research on Optimized Design of Rural Housing in Cold
Regions Based on Parametrization and Machine Learning
Minghui Sun, Yibing Xue * and Lei Wang
Citation: Sun, M.; Xue, Y.; Wang, L.
Research on Optimized Design of
Rural Housing in Cold Regions Based
on Parametrization and Machine
Learning. Sustainability 2024, 16, 667.
https://doi.org/10.3390/su16020667
Academic Editor: Masa Noguchi
Received: 11 November 2023
Revised: 23 December 2023
Accepted: 8 January 2024
Published: 12 January 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Sustainability 2024, 16, 667. https://doi.org/10.3390/su16020667
School of Architecture and Urban Planning, Shandong Jianzhu University, Jinan 250100, China
* Correspondence: xueyb@sdjzu.edu.cn
Abstract: With the rapid development of urbanization, the emergence of more self-built buildings in
the countryside has brought about energy waste problems and decreased comfort. Achieving the
low-carbon goal and improving the quality of the human living environment through architectural
and planning means have become vital issues. In this study, from a parametric perspective, model
building and performance simulation are carried out using Rhino and Grasshopper, and a multi-
objective optimization method and a neural network model are used as the theoretical basis to
train the prediction model after data collection and processing. The model validation of R2 = 0.988
and MSE = 0.0148 indicates that the model can accurately reflect the program’s performance. By
establishing a rapid prediction model for the performance of rural residential buildings, decision-
makers can perform performance predictions under various parameter combinations at the early
design stage, facilitating the screening of building types with high energy consumption and costs. The
method can improve the efficiency of decision-making at the early stage of design, help save decision-
making costs by screening high-energy-consuming building types, improve the living conditions
of residents, reduce carbon emissions, and contribute to the sustainable development of residential
building renewal design in rural areas.
Keywords: rural residential buildings; neural networks; multi-objective optimization; building
energy consumption; sustainable development
1. Introduction
The world faces various problems, such as energy crisis, ecological damage, and
climate warming. The energy used in the building industry accounts for about 40% of the
global energy consumption, and improving the energy efficiency of buildings has positively
impacted the environment and the economy. The IPCC suggests the reduction in the de-
mand for end-of-build energy use from 8% to 14% by 2030 [1], and based on the importance
of this goal, China has proposed a “dual-carbon” strategy and made adjustments to the en-
ergy structure of buildings, in which the energy consumption level of residential buildings
accounts for 21% of the national total energy consumption level. According to the “China
Building Energy Consumption and Carbon Emission Research Report (2022)”, in 2022,
rural buildings consumed 230 million tons of standard coal, accounting for approximately
21% of the total energy consumption of civil buildings in the nation [2]. Due to the lack
of scientific guidance and outdated construction models, energy consumption in rural
buildings remains high. The optimization of energy saving in existing rural dwellings is a
critical aspect for achieving carbon neutrality and peak carbon in the building industry [3].
As China’s economy continues to develop and the national standard of living continues
to improve, the changes associated with it are the production and lifestyles of rural residents
and their behavioral patterns. The focus of most family life functions changes to living.
This change has increased demand for rural residents’ living environment and housing
functions. Nowadays, most newly built farmhouses only focus on appearance and neglect
function, resulting in the conflict between the residential building space and the occupants’
https://www.mdpi.com/journal/sustainability
Sustainability 2024, 16, 667 2 of 19

lifestyle and between the architectural appearance and the aesthetic norms [4]. On the
one hand, this cannot meet the residents’ demand for the use of residential space, and
on the other hand, it also makes the residents increasingly dependent on the indoor air-
conditioning system. Therefore, it is crucial to accelerate the scientific development of rural
residential architecture by scientific renewal and the construction of rural residences.

1.1. Performance Problems of Rural Residential Buildings

Most of China’s rural residential buildings use the millennia-old “Qin brick and Han
tile” construction method, with more than 90% of them consisting of fired clay bricks or
concrete hollow blocks as the main structure [5]. The traditional form of village enclosure
structure has no thermal insulation materials. The thermal insulation performance is poor,
resulting in a high energy consumption in the use phase of the building, and the energy cost
correspondingly increases. Generally speaking, the energy consumption of rural residential
buildings is mainly due to the energy used in daily life, such as cooling, heating, lighting,
cooking, and other energy-intensive electrical equipment [6].
In addition, space heating and cooling are significant contributors to energy-related
carbon emissions [7,8], and heating is the leading cause of energy consumption in rural
buildings in Northern China [9]. To reduce the energy consumption of rural residen-
tial buildings, many scholars have provided some adjustment suggestions for building
planning and design, enclosure structure, renewable energy utilization, and the analysis
of occupancy behavior [10]. In contrast, people continue paying attention to the indoor
thermal environment of rural buildings. For example, Jermyn D. et al. [11] conducted
comprehensive optimization designs on typical residential models, achieving the most
significant optimization effects in the building envelope and heating equipment, with the
highest energy savings reaching up to 80%. Zhang Yi et al. [12] analyzed the thermal envi-
ronment of the traditional farmhouse and the modern farmhouse in the northern part of
Jiangsu Province by comparing and contrasting the traditional farmhouse with the modern
farmhouse and found that the traditional farmhouse enclosure structure has better thermal
stability and has a more significant advantage in terms of thermal insulation in the summer.
Asadi E. et al. [13] proposed that the optimization of building envelope structures should
pursue the best economic efficiency and a lower energy consumption while fulfilling the
residents’ living requirements. The studies mentioned above demonstrate that, owing
to a deficiency in scientific design, numerous rural buildings manifest elevated levels of
energy consumption (EC); however, the indoor thermal environment needs to attain the
anticipated condition.
Due to the large per capita building area in the northern rural areas, residents have
a high degree of freedom to use a particular room or outdoor patio, terrace, and other
spaces. The research found that the residents use the farmhouse space for periods [14]. The
renewal design of rural residential buildings should fully consider the behavioral habits
of residents and incorporate the comfort of the courtyard into the overall consideration.
Relevant studies have shown that when it is necessary to turn on the air-conditioning
indoors to improve the thermal environment, there are some periods when the thermal
environment of the external space is still more pleasant, and residents may choose to
stay outdoors [15–17]. Therefore, improving the comfort of courtyards has an inextricable
impact on reducing the energy consumption of rural residential buildings. This study
aims to understand the current situation of rural residential buildings, provide solutions
and suggestions for improving the comfort of rural residential building complexes, reduce
environmental pollution caused by heating and cooling energy consumption, and provide
the sustainable development of rural residential buildings.

1.2. Application of Optimization Methods

There are many influencing factors for building energy consumption, including build-
ing form parameters, envelope parameters, and shading measures, which have a nonlinear
coupling effect on building energy consumption. The main task of the energy-saving
Sustainability 2024, 16, 667 3 of 19

retrofit of buildings is to find the best combination of energy-saving measures among many
strategies. At the same time, energy-saving retrofit measures interact, meaning that a partic-
ular measure may significantly affect another measure or even achieve the opposite effect.
Thus, simply and linearly adding up the effects of individual measures is not adequately
accurate; hence, the decision-maker will face a multi-objective optimization problem, which
describes multiple competing objective functions to evaluate feasible alternatives [18].
With the continuous development of computer technology, an increasing number of
scholars have investigated the application of multi-objective optimization algorithms in
building energy efficiency. Some researchers focus on studying the optimization problem
regarding computational and building design parameters. For example, Gao Yuan et al. [19]
combined the SPEA-2 optimization algorithm in Rhino-Grasshopper with Energy Plus to
carry out multi-objective optimization design for multiple parameters such as orientation,
depth, window-to-wall ratio, etc., for country houses. Vukadinović et al. [20] used Design
Builder software (2014) to optimize the window-to-wall ratio of passive houses and glass
type, and other parameters were optimized and iterated using Latin hypercube sampling
to determine the heating and cooling energy consumption of the two design scenarios.
In addition, numerous studies on parameters such as building envelope and equipment
systems exist. For example, Fabrizio et al. [21] optimized thermal performance parameters
such as wall thermal inertness, insulation thickness, and external window transmittance
for a Mediterranean residential building envelope with the objectives of primary energy
consumption and thermal discomfort hours. Carlucci et al. [22] combined the GenOpt
optimization engine with the Energy Plus simulation engine to perform a multi-objective
optimization of a net-zero-energy Southern Italian residence for the multi-objective opti-
mization of thermal discomfort in winter and summer and visual discomfort from glare
and the amount of daylight.

1.3. Introduction of Machine Learning

However, traditional methods for the multi-objective optimization of building perfor-
mance often require lengthy computation times and repeated adjustments of each physical
building parameter. Various machine learning algorithms have been used in architecture
recently, including support vector machines (SVM), random forests, artificial neural net-
works (ANN), and many others. In building performance optimization, machine learning
algorithms, as a powerful tool, have been widely used in building design, energy con-
sumption prediction, and comfort assessment. For example, Yan HA et al. [23] integrated
indoor and outdoor environmental data of an office building and used genetic algorithms
and XG Boost algorithm for optimization and prediction in the early design phase of the
office building. Marijana Zekic [24] et al. compared an artificial neural network with
two types of tree-structured regression models and tested three strategies based on the
combination of these three models used to predict the effects, confirming the potential of
hybrid machine learning methods. Guo J et al. [25] used a GANN optimization process
to complete the performance optimization of village houses in the cold regions of China
using building energy consumption, UID, comfort time ratio, and energy saving cost as
optimization objectives.
The paper of Chen-Liang Liu [26] mentions that when training the relationship be-
tween power load and weather data, the hyperparameter c of auxiliary weights is intro-
duced to fit the auxiliary regression, which turns the single-task regression into a multi-task
one. The regression effect of the model is optimized by 21% compared with the single
regression. This suggests that in the regression process of MLP, multiple objectives interact
with each other and improve the regression fitting effect.
Through the compilation of related literature reviews, there are many studies on
optimizing energy savings, thermal environment, and daylighting through building design.
However, there need to be more studies on the optimization of the design of village
residential buildings by comprehensively considering indoor and outdoor environments
and performance. Thus, taking a rural residence in the Jinan area as a case study, this
Sustainability 2024, 16, 667 4 of 19

paper investigates the impact of architectural design parameters and envelope structure
parameters on the residence’s energy efficiency, comfort, and economic viability. It proposes
a method for the rapid prediction of building performance. By integrating MOO (multi-
objective optimization) and ML (machine learning), this method establishes a predictive
model for forecasting multiple objectives. It helps architects predict the performance of
rural homes during retrofits and upgrades, provides decision support in the early stages
of building design, and reduces decision costs. By enhancing the economic benefits of
energy-efficient design as well as the living experience of residents, the goal of sustainable
development is achieved in both economic and social dimensions.

1.4. Purpose of the Study

In general, as the economy grows, rural production and lifestyle, as well as the
behavioral patterns of the residents, change. The demand for comfortable indoor and
outdoor environments in rural residential buildings also increases energy consumption. In
addition, the pursuit of energy-saving measures without considering the cost also leads
to measures that are not possible to implement. Therefore, this study aims to understand
the current situation of rural residential buildings in cold regions, to promote the scientific
design and renewal of rural residential buildings, to improve the performance of rural
residential buildings in terms of energy consumption, comfort, and economy, and to provide
Sustainability 2024, 16, x FOR PEER REVIEW 5 of 19
solutions and effective renewal strategies to guide the design of future rural residential
buildings in cold regions (Figure 1).

Figure
Figure 1.
1. Overall
Overallworkflow.
workflow.

2. Research Methodology and Modeling

2.1. Field Study of Rural Residential Buildings
2.1.1. Cold Regions of China
Most of the cold regions in China are located in the north, necessitating the consider-
ation of both winter and summer insulation objectives. The area of focus for this study is
Jinan City in Shandong Province, situated in the mid-latitude zone and characterized by
a temperate monsoon climate. This climate is distinguished by clear monsoons and dis-
tinct seasons: a dry and rain-scarce spring, a warm and rainy summer, a cool and dry
Sustainability 2024, 16, 667 5 of 19

2. Research Methodology and Modeling

2.1. Field Study of Rural Residential Buildings
2.1.1. Cold Regions of China
Most of the cold regions in China are located in the north, necessitating the considera-
tion of both winter and summer insulation objectives. The area of focus for this study is
Jinan City in Shandong Province, situated in the mid-latitude zone and characterized by a
temperate monsoon climate. This climate is distinguished by clear monsoons and distinct
seasons: a dry and rain-scarce spring, a warm and rainy summer, a cool and dry autumn,
and a cold winter with little snow.

2.1.2. Field Research and Measurement

In this paper, we chose as the object of study the courtyard-style residential buildings
in Dongquan Village (Figure 2), which are situated on Caishi Street, Lixia District, Jinan
City, and have more than 700 years of history and culture. Most of the houses in the village
are in the form of a compound, with the houses facing south, and the south direction is
used chiefly for planting, placing agricultural tools, or as a drying ground. The courtyard
walls are generally tall and thick, effectively dividing the space between inside and outside
the courtyard, which is peaceful, quiet, closed, and safe. Most of the courtyards have east
and west compartments, and people can use the roofs of these compartments for drying
and other purposes. In terms of morphological features, due to historical and geographical
factors, most of the buildings in Dongquan Village adopt the “one house, one courtyard”
model, with the houses in the courtyard adopting the layout of the main house sitting north
to south and constructing compartments according to the needs of the family. The middle
Sustainability 2024, 16, x FOR PEER REVIEW 6 of
of the courtyard is generally spacious, reflecting the centripetal nature of the courtyard and19
strengthening the sense of privacy and belonging.

Figure2.2. Aerial
Figure Aerial view
view of
of the
thelayout
layoutof
ofDongquan
DongquanVillage.
Village.

In order
In order to to improve
improve thethe efficiency
efficiency and and accuracy
accuracy of of the
the field
field survey,
survey, the
the village
village layout
layout
andthe
and theform
formofof building
building layout
layout were
were firstfirst analyzed
analyzed basedbased
on theon the current
current status ofstatus of the
the village
village
plan. In plan. In thehouses,
the village villagethehouses, the main
main house mainlyhouse
has mainly has the
the primary primary
function function
of living as the of
livingfunctional
main as the main functional
space; space; itsform
its architectural architectural form
is a regular is a regular
rectangle; rectangle; the com-
the compartments and
partments
inverted and inverted
houses mainly havehouses mainly
storage andhave storage
other and functional
auxiliary other auxiliary functional
spaces. spaces.
By integrating
By integrating
the the existing
existing village village
layout plans andlayout plans and
conducting conducting
actual surveyingactual surveying
in Dongquan in Dong-
Village, we
quan Village,
collected we collected
dimensional dimensional
and structural dataand
forstructural
various typesdataofforbuilding
variouscourtyards.
types of building
After
courtyards.
assessing theAfter assessing
research data of the research
183 data of
farmhouses in 183 farmhouses
the village, in thethat
we know village, we know
the houses in
that the houses in the village are mainly categorized into “—”, “L”, “U”, and “□” shapes.
the village are mainly categorized into “—”, “L”, “U”, and “ □ ” shapes. The four types of
The four
houses intypes of houses
the village in the“L”,
are “—”, village
“U”,are “—”,
and “□”,“L”,
as “U”,
shown andin“□”, as shown
Figure 3. The in Figureof
number 3.
The numberand
“U”-shaped of “U”-shaped
“Mouth”-shaped and “Mouth”-shaped dwellings
dwellings is the highest, is the highest,
accounting accounting
for 36.1% and 34.4%,for
36.1% and 34.4%,
respectively, followed respectively,
by “L”-shaped followed by “L”-shaped
dwellings dwellings atdwellings
at 23%. The “—”-shaped 23%. Theare “—”-
the
shaped dwellings are the least, only 6.6%. Therefore, the objectives of this study are mainly
least, only 6.6%. Therefore, the objectives of this study are mainly the following:
the following:
(1) Type 2: “L”-shaped courtyard house.
(2) Type 3: “U”-shaped courtyard house.
(3) Type 4: “□”-shaped courtyard house.
 courtyards. After assessing the research data of 183 farmhouses in the village, we know
that the houses in the village are mainly categorized into “—”, “L”, “U”, and “□” shapes.
The four types of houses in the village are “—”, “L”, “U”, and “□”, as shown in Figure 3.
The number of “U”-shaped and “Mouth”-shaped dwellings is the highest, accounting for
Sustainability 2024, 16, 667
36.1% and 34.4%, respectively, followed by “L”-shaped dwellings at 23%. The 6“—”- of 19
shaped dwellings are the least, only 6.6%. Therefore, the objectives of this study are mainly
the following:
(1) Type 2: “L”-shaped courtyard house.
(1) Type 2: “L”-shaped courtyard house.
(2) Type 3: “U”-shaped courtyard house.
(2) Type 3: “U”-shaped courtyard house.
(3) Type 4: “□”-shaped courtyard house.
(3) Type 4: “□”-shaped courtyard house.

Figure 3.
Figure 3. Schematic
Schematic diagram
diagram of
of rural
rural residential
residential building
building forms.
forms.

An analysis
An analysis ofof survey
surveydata
datahas
hasrevealed
revealedthat
thatthethescale
scaleofofcourtyards
courtyards is is primarily
primarily de-
deter-
termined
mined by by
thethe dimensions
dimensions of the
of the mainmain house
house andandthe the
wing wing rooms,
rooms, which,
which, together
together withwith
the
the surrounding walls, form the courtyard. The variation in width primarily
surrounding walls, form the courtyard. The variation in width primarily reflects differences reflects dif-
ferences
in in the functional
the internal internal functional layout.
layout. With theWith
rise the rise in economic
in economic levels
levels and and increasing
increasing usage
usage demands,
demands, enhancing
enhancing the building’s
the building’s widthwidth to increase
to increase its area
its floor floorallows
area allows for addi-
for additional
tional functionalities,
functionalities, such assuch
the as the provision
provision of garages
of garages or the or the layout
layout of primary
of primary and sec-
and secondary
ondary bedrooms.
Sustainability 2024, 16, x FOR PEER REVIEW
bedrooms. Similarly,
Similarly, the differences
the differences in depth in also
depth also reflect
reflect variations
variations in the in the internal
7 of
internal 19
func-
functional
tional layoutlayout ofhouse.
of the the house. Buildings
Buildings withwith
lowerlower economic
economic status
status andand earlier
earlier construc-
construction
tion dates
dates typically
typically have have
moremore singular
singular functions,
functions, merely merely
catering catering to cooking
to cooking and needs,
and living living
needs,
resultingresulting
in in relatively
relatively more more
minor minor
overalloverall
depths. depths.
As As economic
economic
the era of house construction changes, the functions within houses gradually become levels
levels improve
improve and and
the
era of house construction changes, the functions within houses
more numerous and diverse, leading to an increase in the depth of the buildings. gradually become more
numerous and diverse, leading to an increase in the depth of the buildings.
To create the surveyed cohort types of the representatives of most rural residential
To create the surveyed cohort types of the representatives of most rural residential
buildings in the village, the range of values taken from six perspectives, “Principal House
buildings in the village, the range of values taken from six perspectives, “Principal House
Width” (PHW), “Principal House Depth” (PHD), “Wing House Width” (WHW), “Wing
Width” (PHW), “Principal House Depth” (PHD), “Wing House Width” (WHW), “Wing
House Depth” (WHD), “Length of the Courtyard” (LIC), and “Width of the Courtyard”
House Depth” (WHD), “Length of the Courtyard” (LIC), and “Width of the Courtyard”
(WIC), were used to understand the characteristics of the rural residential buildings. As
(WIC), were used to understand the characteristics of the rural residential buildings. As
shown in Figure 4, Type 2 PHW is mainly in the range of 10–15. Type 4 has a more even
shown in Figure 4, Type 2 PHW is mainly in the range of 10–15. Type 4 has a more even
yard size distribution than the other types. The performance of WHD for all three types
yard size distribution than the other types. The performance of WHD for all three types of
of compounds is more dispersed, indicating that this indicator is subject to more signifi-
compounds is more dispersed, indicating that this indicator is subject to more significant
cant variation depending on the type of compound. The above information on PHW,
variation depending on the type of compound. The above information on PHW, PHD,
PHD, WHW, WHD, LIC, and WIC for the different compound types was obtained by
WHW, WHD, LIC, and WIC for the different compound types was obtained by compiling
compiling the data
the data from from theresearch
the building buildingin research in the village.
the village.

Figure
Figure4.4.Range
Rangeofofvalues
valuesofofcritical
criticalvariables
variablesfor
fordifferent
differenttypes
typesofofbuildings.
buildings.

3. The Research Process

3.1. Performance Simulation
3.1.1. Optimization Objective Selection
This study focuses on improving the building energy consumption, the thermal com-
fort of the compound, and the building envelope cost. In order to assess the performance
of the building, this study selects Total Load Intensity (TLI), Cost of Envelope Structure
Sustainability 2024, 16, 667 7 of 19

3. The Research Process

3.1. Performance Simulation
3.1.1. Optimization Objective Selection
This study focuses on improving the building energy consumption, the thermal com-
fort of the compound, and the building envelope cost. In order to assess the performance
of the building, this study selects Total Load Intensity (TLI), Cost of Envelope Structure
(CES), and Universal Thermal Climate Index (UTCI) as performance evaluation metrics.
Total Load Intensity (TLI) is valuable for evaluating a building’s annual energy effi-
ciency [27]. This metric is determined by dividing the building’s yearly energy consumption
by its floor area. A reduced TLI indicates greater energy efficiency and diminished carbon
emissions. During the initial design phases, this metric evaluates a building’s energy
efficiency or compares various buildings, providing essential guidance for architects.
In the design of rural residential buildings, while considering the energy efficiency and
the carbon reduction of buildings, it is also necessary to consider the cost factors of building
design to promote and ensure the practical feasibility of the buildings on the ground. Since
this study focuses on the impact of geometric parameters of rural residential buildings on
the objectives, in the current Chinese market cost calculations, only the insulation costs
of the envelope structures (external walls and roofs) affected by these parameters and
the costs of different types of windows are considered. The specific formula for the cost
calculation is as follows:

Cost = Costwall + Costroo f + Costwindow (1)

The meaning of terms present in Equation (1):

Cost = the total cost of the envelope;
Cwall = the cost of building envelope insulation;
Croo f = the cost of building roof insulation;
Cwindow = the cost of windows.
For the thermal environment of the courtyard space, this paper usually adopts the
Universal Thermal Climate Index (UTCI) for calculation, which describes the human
perception of ambient temperature and humidity. It provides a more comprehensive and
accurate measure of human comfort and responds to the human body’s ability to adapt
to the environment than traditional temperature indices [28]. In this study, we use the
Universal Thermal Climate Index to assess the comfort level of the external environment
of a building and the impact of the environment on the building. The Universal Thermal
Climate Index can cope with changes in the building’s outdoor thermal environment due
to changes in yard space and building form.

3.1.2. Parameterization
The parametric model created with Rhino and Ladybug Tools was able to simulate
the performance of the building based on the local climate and the site environment. HB
Energy is used to simulate the energy consumption of the building by using Ladybug from
https://www.ladybug.tools/epwmap/ (accessed on 23 October 2023) to download the
Jinan City weather file as the weather file for this simulation study.
The parameters for the energy simulation were set as follows: the simulation period was
the whole year, the heat transfer coefficient of the facade was taken as K = 0.959 (W·m−2 ·K−1),
the heat transfer coefficient of the roof was taken as K = 1.167 (W·m−2 ·K−1), and the type
of windows was taken as Type1. The personnel presence rate, lighting power, and electrical
equipment power are set according to national and local standards.

3.1.3. Parameterized Model Construction

In this study, the Walleci plug-in in Grasshopper was used to carry out optimization
iterations through the NSGA-II multi-objective genetic algorithm it carries, to obtain the
dataset containing the values of design variables and optimization objectives, and to derive
Sustainability 2024, 16, 667 8 of 19

Sustainability 2024, 16, x FOR PEER REVIEW

the Pareto frontier solution and the non-frontier solution. In this way, more design solution 8 of 19
data are generated and integrated to improve the applicability and accuracy of the machine
learning model. Based on the previous research data for residential buildings within
Dongquan
𝐶 = the cost of
Village, windows.
Type 2, Type 3, and Type 4, which account for a high percentage of the
total number of buildings, are taken as the objects of research, of which Type 2 has two
For the thermal environment of the courtyard space, this paper usually adopts the
forms of arrangement. Table 1 shows the variable system, and Figure 5 shows the schematic
Universal Thermal Climate Index (UTCI) for calculation, which describes the human per-
diagram of the four-parameter models.
ception of ambient temperature and humidity. It provides a more comprehensive and ac-
curate measure of human comfort and responds to the human body’s ability to adapt to
Table 1. Variable setting and initial investment (definition of abbreviations: PHW = principal house
the environment than traditional temperature indices [28]. In this study, we use the Uni-
width; PHD = principal house depth; WHW (W, E) = wing house width (western, east); WHD = wing
versal Thermal Climate Index to assess the comfort level of the external environment of a
house depth; RS = roof slope; EW = eave width; HCW = height of the courtyard’s wall; BH = building
building and the impact of the environment on the building. The Universal Thermal Cli-
height; BO = building orientation; WT = window type; and EPS, XPS = insulation material type).
mate Index can cope with changes in the building’s outdoor thermal environment due to
changes
BuildinginParameters
yard space and building
Range ofform.
Values Unit Initial Investment
PHW 10–20 m /
3.1.2. Parameterization
PHD 3–12 m /
The
WHW parametric
(W, E) model created3–8 with Rhino and Ladybug m Tools was able to / simulate
the performance of the building based on the local climate ◦and the site environment. HB
WHD 3–10 m /
RS 0, 15, 30, 45 /
Energy is used to simulate the energy consumption of the building by using Ladybug
EW 0–0.8 m /
from h ps://www.ladybug.tools/epwmap/
HCW 1.5–3 (accessed on 23 October
m 2023) to download
/ the
Jinan CityBHweather file as the weather 3.0–4.5file for this simulationm study. /
The parameters
BO for the energy simulation were set as follows:
−15–15 ◦ the simulation
/ period
was the
WT whole
(Table 2)year, the heat1;transfer
2; 3; 4; 5;coefficient
6; 7 of the facade
/ was taken as/K = 0.959
−3 )
(W·m ·K ), the heat transfer coefficient of the roof was taken as K = 1.167240
−2 EPS
−1 0.03; 0.05; 0.07; 0.09; 0.11; 0.13 m (W·m −2·K−1), and
(CNY/m
XPS 0.02; 0.04; 0.06; 0.08; 0.1; 0.12 m 260 (CNY/m−3 )
the type of windows was taken as Type1. The personnel presence rate, lighting power,
and electrical equipment power are set according to national and local standards.
Table 2. Types of exterior windows (definition of abbreviations: Low-E = low emissivity coated glass;
3.1.3. Parameterized
T = transmission; Model =Construction
and number thickness (mm)).
In this study, the Walleci plug-in in Grasshopper was used to carry out optimization
Type iterations
Tectonicthrough
Level the NSGA-II multi-objectiveSHGC genetic K (Walgorithm
·m−2 ·K−1 )it carries, to obtain
Initial Investmentthe
(CNY/m−2 )
dataset containing the values of design variables and optimization objectives, and to de-
1 6clearrive
glassthe Pareto
+ 9Air frontier
+ 6clear glass solution and the0.752
non-frontier solution.
3.12 In this way, more 192 design
2 6clear solution data are generated and integrated to improve the applicability and 241
glass + 12Air + 6clear glass 0.725 2.93 accuracy of
3 6Low-E + 9Air + 6clear glass + 200Air + 6Low-E 0.421 2.35 290
the machine learning model. Based on the previous research data for residential buildings
4 6Low-E + 9Air + 6Low-E (High T) 0.477 2.40 265
5 6Low-E within
+ 9AirDongquan
+ 6Low-E (LowVillage,
T) Type 2, Type 3, and Type 4, which
0.453 2.11 account for a high282 percent-
6 6clear glass + 9Airage of theglass
+ 6clear total+ number
200Air + of buildings,
6clear glass are 0.532
taken as the objects
2.34of research, of which
260 Type 2
7 has two
3Low-E forms
+ 0.1Air of arrangement. Table 1 shows
+ 3Low-E 0.656 the variable 2.22
system, and Figure 240 5 shows the
schematic diagram of the four-parameter models.

Figure 5. Four types of parameterized models.

Figure 5. Four types of parameterized models.

Table 1. VariableAnalysis
3.2. Correlation se ing and initial investment (definition of abbreviations: PHW = principal house
width; PHD = principal house depth; WHW (W, E) = wing house width (western, east); WHD = wing
This section discusses the architectural parameter variables and simulates the relevant
house depth; RS = roof slope; EW = eave width; HCW = height of the courtyard’s wall; BH = building
parameter variables of the main house and compartments, respectively. Subsequently, a
height; BO = building orientation; WT = window type; and EPS, XPS = insulation material type).
Pearson correlation analysis of the data was conducted using the “scipy.stats” library in
Building Parameters Range of Values Unit Initial Investment
PHW 10–20 m /
PHD 3–12 m /
 7 3Low-E + 0.1Air + 3Low-E 0.656 2.22 240

3.2. Correlation Analysis

Sustainability 2024, 16, 667

This section discusses the architectural parameter variables and simulates the rele-
9 of 19
vant parameter variables of the main house and compartments, respectively. Subse-
quently, a Pearson correlation analysis of the data was conducted using the “scipy.stats”
library in Python. If the p-value is less than 0.05, it indicates a significant correlation be-
Python. If the p-value is less than 0.05, it indicates a significant correlation between the two
tween the two variables. A correlation coefficient close to 1 implies a strong positive cor-
variables. A correlation coefficient close to 1 implies a strong positive correlation between
relation between the variables, while a coefficient close to −1 indicates a strong negative
the variables, while a coefficient close to −1 indicates a strong negative correlation. Typical
correlation. Typical reference models have the main house with a width of 16 m, a depth
reference models have the main house with a width of 16 m, a depth of 6 m, and a roof
of 6 m,ofand
slope 30◦ ,aand
roofthe
slope of 30°, and with
compartments the compartments
a width of 3 mwith
and aa depth
widthof
of63m.
m Each
and aparameter
depth of
6study
m. Each
controls a single variable for simulation, and the specific simulation processsim-
parameter study controls a single variable for simulation, and the specific and
ulation process and results
results are as follows (Figure 6).are as follows (Figure 6).

Sustainability 2024, 16, x FOR PEER REVIEW 10 of 19

Figure 6. Simulation results of building parameters.

Figure 6. Simulation results of building parameters.
From the line graph, PHW, PHD, WHW, WHD, and BH have a significant correlation
with From
TLI, COST, and
the line UTCI,
graph, and PHD,
PHW, RS hasWHW,
a significant
WHD, andcorrelation
BH have with TLI, and there
a significant is no
correlation
significant correlation
with TLI, COST, with COST
and UTCI, and RSand
hasUTCI.
a significant correlation with TLI, and there is no
significant correlation with COST and UTCI.
3.2.1. Simulation Analysis of Courtyard Parameters
3.2.1.InSimulation Analysis
this section, of Courtyard
the courtyard Parameters
parameter variables are explored and simulated sepa-
ratelyInfor thesection,
this height of
thethe courtyard
courtyard wall, orientation,
parameter variablesand
arethe width of
explored thesimulated
and eaves. Among
sepa-
them,
rately the
for orientation
the height ofisthe
represented
courtyardby 1–7orientation,
wall, for values of
and−15°–15° change
the width (ineaves.
of the 5° intervals).
Among
◦ –15◦ change (in 5◦ intervals).
Each
them,parameter studyiswas
the orientation simulatedby
represented by1–7
controlling only
for values of −a 15
single variable, and the specific
simulation
Each parameterprocess andwas
study results are as follows
simulated (Figureonly
by controlling 7). a single variable, and the specific
simulation process and results are as follows (Figure 7).

Figure
Figure 7.
7. Simulation
Simulation results
results of
of courtyard
courtyard parameters.
parameters.

The line
The line graph
graphshows
showsthat
thatBO, EW,
BO, andand
EW, HCW werewere
HCW significantly correlated
significantly with UTCI
correlated with
and not significantly correlated with TLI and COST.
UTCI and not significantly correlated with TLI and COST.

3.2.2. Simulation and Analysis of Building Envelope Parameters

This section investigates the building envelope insulation, roof insulation thickness,
and window type, and then simulates the values of the target variables under the control
of the three parameters. Typical reference models for the envelope insulation, roof insula-
 Figure 7. Simulation results of courtyard parameters.

The line graph shows that BO, EW, and HCW were significantly correlated with
Sustainability 2024, 16, 667 UTCI and not significantly correlated with TLI and COST. 10 of 19

3.2.2. Simulation and Analysis of Building Envelope Parameters

3.2.2.This section and
Simulation investigates
Analysisthe building Envelope
of Building envelope insulation,
Parametersroof insulation thickness,
and window type,investigates
This section and then simulates the values
the building envelopeof the target variables
insulation, under the
roof insulation control
thickness,
of
andthe three parameters.
window Typical
type, and then reference
simulates models
the values offor
thethe envelope
target insulation,
variables roof
under the insula-
control of
tion thickness,
the three and exterior
parameters. Typicalwindow
referencetype are for
models set the
to 0.03, 0.02, insulation,
envelope and 1, respectively, with
roof insulation
each parameter
thickness, controlling
and exterior only one
window typevariable
are set for the study
to 0.03, 0.02, (Figure 8).
and 1, respectively, with each
parameter controlling only one variable for the study (Figure 8).

Figure
Figure 8.
8. Simulation
Simulation results
results of
of enclosure structure parameters.
enclosure structure parameters.

As shown
As shown inin the
the line
line graphs,
graphs, EPS
EPS and
and XPS
XPS have
have aa significant
significant correlation
correlation with
with TLI
TLI and
and
COST but
COST but no
no significant
significant correlation
correlation with
with UTCI.
UTCI. WT
WT has
has no
no significant
significant correlation
correlation with
with all
all
three but has a varying degree of influence on TLI and
three but has a varying degree of influence on TLI and COST.COST.

3.2.3. Factor Correlation Analysis

3.2.3. Factor Correlation Analysis
Based on the above factor correlation analysis, building parameters have the most
Based on the above factor correlation analysis, building parameters have the most
significant influence on TLI. Changes in the design values of the building parameters can
significant influence
directly affect onofTLI.
the TLI the Changes
building.inNoteworthily,
the design values of the building
the correlation parameters
between COST andcan
the building parameters is also crucial. People must consider design, TLI, and COST in
the building process. Too large or too small building parameters will increase the energy
consumption of the building. From the analysis, some building parameters also correlate
highly with the UTCI of the courtyard space, such as WHW. It mainly shades part of the
sunlight projected into the courtyard, which makes the solar radiation received in the
courtyard lower, resulting in a lower UTCI.
The courtyard parameters all significantly affect the UTCI, while they do not signifi-
cantly affect TLI and COST because they are not a part of the building. However, it is worth
noting that the BO has no significant effect on the TLI, but it does influence the TLI to a
certain extent due to its change in orientation, which affects the light and heat gain from
the windows of the building.
The envelope parameters have a direct impact on TLI and COST. With the increase in
the thickness of the insulation layer of the envelope, the TLI gradually decreases. However,
its decreasing trend gradually slows down, which indicates that the increase in the thickness
of the insulation layer does not achieve the same effect, while the COST is steadily increas-
ing. TLI and COST should be taken into account when selecting envelope parameters.
Otherwise, there will be wasted performance.

3.3. Multi-Objective Optimization Results

Multi-Objective Optimization
The research utilized Wallacei [28] as its primary tool for the multi-objective optimiza-
tion. This tool facilitates evolutionary simulations within Grasshopper, leveraging genetic
algorithms alongside the Pareto frontier theory. It aids designers in comprehending the
nuances of the evolutionary selection process by offering intricate analytical tools and
cohesive selection strategies. Optimization is crucial at every phase of the evolutionary
simulation process, ensuring thorough and efficient analysis.
Total Load Intensity (TLI), Envelope Cost, and Universal Thermal Climate Index
(UTCI) were entered into the target port as optimization objectives, where UTCI calculates
Sustainability 2024, 16, 667 11 of 19

the mean. There may be slight deviations in the evaluation results, but they have little
effect on the trend of the optimization results. Then, the variables of building parameters,
as well as the initial investment in Tables 3 and 4, are input into the gene port to optimize
Type2, Type 3, and Type 4, respectively. The number of iterations is 40 generations, the
subset of each generation is 50, and the number of populations is 2000. Table 3 shows the
genetic algorithm parameters.

Table 3. Genetic algorithm parameter settings.

Setting Parameters Numerical Value Type

Crossover Probability 0.9 Ratio
Mutation Probability 1/r Ratio
Crossover Distribution Index 20 Number
Mutation Distribution Index 20 Number
Random Seed 1 Number

Table 4. Random case parameter variables.

RF PHW PHD WWHW EWHW WHD BH EW HCW BO EPS XPS

Type WT
(m) (m) (m) (m) (m) (m) (m) (m) (m) (◦ ) (m) (m)
1 2 1 17 4 6 4 10 3.1 0 1.5 −11 0.07 0.07 0
2 2 2024, 16,
Sustainability 1 x FOR 14 3
PEER REVIEW 5 7 7 3 0.1 1.5 −7 0.1 0.1 12 of
1 19
3 2 0 14 8 5 8 7 3 0.2 1.6 −9 0.11 0.12 0
4 3 3 18 8 6 4 6 3.5 0.5 2.2 −9 0.1 0.12 1
5 3 2 14 5 3 7 6 3 0.2 2.5 4 0.11 0.12 0
66 3 11 19
19 5 5 3 3 3 3 3 3 3 3 0 0 1.5 1.5 −7 −7 0.10.1 0.11
0.11 0
77 44 11 15
15 5 5 7 7 4 4 4 4 3.8 3.8 0.6 0.6 2.8 2.8 −5 −5 0.060.06 0.07
0.07 55
88 44 00 19
19 5 5 7 7 4 4 8 8 3 3 0.2 0.2 2.6 2.6 −3 − 3 0.13
0.13 0.09
0.09 00
9 4 2 18 10 5 8 7 3 0.4 2.6 1 0.12 0.12 6
9 4 2 18 10 5 8 7 3 0.4 2.6 1 0.12 0.12 6

Figure 99shows
Figure showsthethesolution
solutionset
setofofthe
the multi-objective
multi-objective optimization
optimization as aasscatter
a sca plot
er plot
on
on the Wallacei coordinate system. TLI, Cost, and UTCI correspond to the x, y,
the Wallacei coordinate system. TLI, Cost, and UTCI correspond to the x, y, and z axes. and z axes.
The
The image
PCP PCP image in Figure
in Figure 10 the
10 shows shows the analysis
analysis of the optimization
of the optimization process toprocess
achieve to achieve
minimum
minimum
TLI, TLI, COST,
minimum minimumand COST,
maximum and UTCI.
maximum UTCI.

Figure 9.
Figure 9. Distribution
Distribution of
of Pareto
Pareto frontier
frontier solutions
solutions for
for three
three types
types of
of RRB.
RRB.

Figure 10. Distribution of parallel coordinate plot for three types of RRB.

By analyzing the results of the multi-objective optimization and the standard devia-
 Sustainability 2024, 16, 667 12 of 19
Figure 9. Distribution of Pareto frontier solutions for three types of RRB.

Figure
Figure10.
10.Distribution
Distributionof
ofparallel
parallelcoordinate
coordinateplot
plotfor
forthree
threetypes
typesof
ofRRB.
RRB.

Byanalyzing
By analyzing the
the results
results of
of the
themulti-objective
multi-objectiveoptimization
optimizationandandthe standard
the standard deviation
devia-
plots of the process, the overall evolutionary direction of the algorithm tends
tion plots of the process, the overall evolutionary direction of the algorithm tends to be stable,
to be
and all of them converge to the optimal trend. Among the 194 Pareto frontier solutions of
stable, and all of them converge to the optimal trend. Among the 194 Pareto frontier solu-
Type 2, the objective range of COST is from 8000 to 28,000, the objective range of TLI is from
tions of Type 2, the objective range of COST is from 8000 to 28,000, the objective range of
70 kWh/m2 to 148 kWh/m2 , and the mean value of UTCI is between 16.7 ◦ C and 17.3 ◦ C;
TLI is from 70 kWh/m2 to 148 kWh/m2, and the mean value of UTCI is between 16.7 °C
among the 229 Pareto frontier solutions of Type 3, the objective range of cost is from 9000 to
and 17.3 °C; among the 229 Pareto frontier solutions of Type2 3, the objective range of cost
32,000, the target range of TLI is 70 kWh/m2 to 160 kWh/m , and the mean value of UTCI
is from 9000 to 32,000, the target range of TLI is 70 kWh/m 2 to 160 kWh/m2, and the mean
is between 16.2 ◦ C and 17.2 ◦ C. For the 326 Pareto frontier solutions of Type 4, the target
value of UTCI is between 16.2 °C and 17.2 °C. For the 326 Pareto frontier solutions of Type
range of cost is 13,000 to 48,000, the target range of TLI is 63 kWh/m2 to 138 kWh/m2 , and
4, the target range of cost is 13,000 to 48,000,◦the target◦range of TLI is 63 kWh/m2 to 138
the mean value of UTCI ranges between 16 C and 17 C. The average value of the
Sustainability 2024, 16, x FOR PEER REVIEW
UTCI
13 value
of 19
kWh/m2, and the mean value of UTCI ranges between 16 °C and 17 °C. The average
is between 16 ◦ C and 17 ◦ C. The average value of the UTCI is between 16 ◦ C and 17 ◦ C
of the UTCI is between 16 °C and 17 °C. The average value of the UTCI is between 16 °C
(Figure 11).
and 17 °C (Figure 11).

Figure
Figure11.
11.Box
Boxdiagram
diagramofofPareto
Paretofrontier
frontiersolution
solutiondistribution.
distribution.

The
Theabove
abovedata
datawere
wereexported
exportedfrom
fromGrasshopper
GrasshoppertotoExcel
Excelvia
viathe
theTT
TTToolbox
Toolboxtool
toolfor
for
the
thefollowing
followingmachine
machinelearning
learningmodel
modeltraining.
training.

3.4.Predictive
3.4. PredictiveModel
ModelConstruction
Construction
Thisstudy
This studyemploys
employsthe thefeedforward
feedforwardneural
neuralnetwork
networkalgorithm
algorithmand andcommonly
commonlyused used
machine learning libraries for model construction. Scikit-learn is the premier Python library
machine learning libraries for model construction. Scikit-learn is the premier Python li-
for machine learning, constructed upon foundational data science packages like numpy
brary for machine learning, constructed upon foundational data science packages like
and matplotlib. It encompasses various machine learning functionalities, including data
numpy and matplotlib. It encompasses various machine learning functionalities, includ-
sampling, preprocessing, model validation, classification, and regression [29]. In this study,
ing data sampling, preprocessing, model validation, classification, and regression [29]. In
a regression prediction model is employed. Neural networks, which are artificial neural
this study, a regression prediction model is employed. Neural networks, which are artifi-
networks inspired by biological neural systems, are applicable in various machine learning
cial neural networks inspired by biological neural systems, are applicable in various ma-
applications, including regression prediction. Furthermore, this method is noted for its
chine learning applications, including regression prediction. Furthermore, this method is
comprehensive efficiency and accuracy as a machine learning technique.
noted for its comprehensive efficiency and accuracy as a machine learning technique.
Neural network fitting for regression prediction is a technique for making predictions
Neural network fi ing for regression prediction is a technique for making predictions
on a given dataset by training a neural network. In a neural network, nodes are organized
on a given dataset by training a neural network. In a neural network, nodes are organized
into a hierarchy. The input layer receives raw data, the intermediate layer processes
into
the ainput
hierarchy. The extracts
data and input layer receives
features, andrawthedata, thelayer
output intermediate
generates layer processesEach
predictions. the
input data and extracts features, and the output layer generates predictions.
node has an associated weight and bias, and these values can be trained and adjusted by Each node
has an associated weight
backpropagation and to
algorithms bias, and these
minimize valuesbetween
the error can be trained and adjusted
the predicted and actualby results.
back-
propagation algorithms to minimize the error between the predicted and actual
The fitting process is usually divided into two phases: training and testing. In the training results.
The fi ing
phase, theprocess
datasetisisusually
divided divided into twoand
into training phases: training
validation andThe
sets. testing. In theset
training training
is then
phase,
used tothetrain
dataset
the is divided
neural into training
network to findand
the validation sets. The
optimal weights andtraining
biasessetforisminimizing
then used
to train the neural network to find the optimal weights and biases for minimizing the pre-
diction error. During training, optimization algorithms such as stochastic gradient descent
can help optimize the weights and biases of the neural network.

3.4.1. Data Preprocessing

Sustainability 2024, 16, 667 13 of 19

the prediction error. During training, optimization algorithms such as stochastic gradient
descent can help optimize the weights and biases of the neural network.

3.4.1. Data Preprocessing

Data preprocessing involves identifying, rectifying, or eliminating flawed, inaccurate,
or irrelevant data within a dataset. This process aims to convert raw data into a format
suitable for training, ensuring that the refined data aligns well with the specifications and
needs of the model [30].
To eliminate the differences in the data range between variables and to avoid the
objective function being unable to learn accurately from the data due to significant differ-
ences in feature sizes in the dataset, we process the data in the dataset using the Z-score
normalization method, transforming it into data with a mean of 0 and a standard deviation
of 1. The standardized value of each parameter is calculated as (X − mean)/std, where
“mean” represents the average value, and “std” denotes the standard deviation.

3.4.2. Building and Training Model

This study established a machine learning model using the PyTorch(1.7.0) framework.
Sustainability 2024, 16, x FOR PEER REVIEW 14 of 19
The model was trained with parameters from existing design schemes and simulated target
values, aiming to effectively predict the performance levels of RRB and assist designers in
decision-making. A multilayer feedforward neural network (MLP) based on the backpropa-
response variables.
gation algorithm Thecan
(LM) predictive
be usedmodel, as shown
to perform in Figureof
the regression 12, comprises
multiple an input
response layer,
variables.
two
The hidden layers,
predictive andas
model, anshown
outputinlayer. Each
Figure 12,node corresponds
comprises an inputtolayer,
a response variable
two hidden and
layers,
possesses its unique
and an output layer. weights and
Each node biases, employing
corresponds the “relu”
to a response activation
variable function.
and possesses itsDuring
unique
the training
weights andphase,
biases,the network optimizes
employing the “relu”these weights
activation and biases
function. through
During the backprop-
the training phase,
agation algorithm
the network to minimize
optimizes the error
these weights andbetween predicted
biases through the and actual outputs.
backpropagation This neu-
algorithm to
ral network
minimize theiserror
capable of handling
between predicted complex datasets,
and actual including
outputs. thosenetwork
This neural with multiple re-
is capable
sponse variables.
of handling complex datasets, including those with multiple response variables.

Figure
Figure 12.
12. Structure
Structure diagram
diagram of
of MLP
MLP neural
neural network.
network.

In
In MLP,
MLP, the
the data
data are
are input
input as
as aa matrix,
matrix, and
and neurons
neurons areare activated
activated in
in the
the implicit
implicit layer
layer
for
for weight
weightresponse.
response. TheThemodel
modelcancanconsider
considerthetheeffect
effectof
ofeach
eachperformance
performancefeature
featureduring
during
backpropagation.
backpropagation. The relationship between
The relationship betweenthe theeffects
effectsofofthethethree
three performance
performance parame-
parameters
ters impacts the fi ing process. The data in this paper are fi ed to a neural network
impacts the fitting process. The data in this paper are fitted to a neural network using
using an
an array of 6000 observations with 14 features as predictor variables and observations with
array of 6000 observations with 14 features as predictor variables and observations with
three features
three featuresas asresponse
responsevariables.
variables.The The “Batch
“Batch Size”
Size” was wassetset to 128,
to 128, andand the the model’s
model’s re-
regression performance was evaluated using the “MSE” (Mean Squared
gression performance was evaluated using the “MSE” (Mean Squared Error) loss function Error) loss function
and R2
and R2 (R-Square)
(R-Square) metrics.
metrics.
The dataset is divided into
The dataset is divided into aatraining
trainingsetsetfor
fortraining
trainingthethemachine
machinelearning
learningmodel
modelandand
a test set for validating the model. In this study, 70% of the data was randomly
a test set for validating the model. In this study, 70% of the data was randomly selected as selected as
the training set, 20% as the test set, and 10% as the validation
the training set, 20% as the test set, and 10% as the validation set. set.

3.4.3. Model Evaluation

The training regression was conducted by modeling the neural network through Py-
thon for a total of 300 rounds and converged after the 100th round. Cross-validation was
used to select and evaluate simulations, culminating in choosing the best-performing
model. The specific model effect is shown in Figures 13 and 14. From the fi ing effect
graph, the model has R2 = 0.988 for the training set, R2 = 0.986 for the testing set, R2 = 0.985
for the validation set, and R2 = 0.988 for all data. The prediction model converged in all

Sustainability 2024, 16, 667
three features as response variables. The “Batch Size” was set to 128, and the model’s re-
gression performance was evaluated using the “MSE” (Mean Squared Error) loss function
and R2 (R-Square) metrics.
The dataset is divided into a training set for training the machine learning model and
a test set for validating the model. In this study, 70% of the data was randomly selected as
14 of 19
the training set, 20% as the test set, and 10% as the validation set.

3.4.3. Model
3.4.3. Model Evaluation
Evaluation
The training
The training regression
regressionwaswasconducted
conductedbyby modeling
modeling thethe
neural network
neural networkthrough Py-
through
thon forfor
Python a total of 300
a total rounds
of 300 roundsandandconverged
converged after the the
after 100th round.
100th Cross-validation
round. Cross-validationwas
usedused
was to select andand
to select evaluate
evaluatesimulations,
simulations,culminating
culminatingininchoosing
choosing the best-performing
the best-performing
model. The
model. The specific
specific model effect is
model effect is shown
shown in in Figures
Figures 1313 and
and 14.
14. From
From thethe fitting
fi ing effect
effect
graph, the
graph, the model
modelhas
hasR2R2== 0.988
0.988 for
for the
the training
trainingset,
set,R2
R2== 0.986
0.986 for
for the
the testing
testing set,
set,R2
R2== 0.985
0.985
for the validation set, and R2 = 0.988 for all data. The prediction model converged in all
for the validation set, and R2 = 0.988 for all data. The prediction model converged in all
three parts,
three parts, thus
thus avoiding
avoiding the
the issue
issue ofof overfitting.
overfi ing.

Sustainability 2024, 16, x FOR PEER REVIEW 15 of 19

Figure 13.
Figure 13. Training
Training and
and validation
validation loss
loss over
over time.
time.

Figure 14. Neural network model regression graph.

Figure 14. Neural network model regression graph.

3.4.4. Code Encapsulation

After validating the feedforward neural network prediction model constructed based
on the multi-objective dataset of rural residential buildings in cold regions, the system
code was encapsulated to enhance its usefulness for designers. The software was built
with Tkinter, a standard GUI library for Python, and mainly predicted the three target
Sustainability 2024, 16, 667 15 of 19
Figure 14. Neural network model regression graph.

3.4.4.Code
3.4.4. CodeEncapsulation
Encapsulation
Aftervalidating
After validatingthe
thefeedforward
feedforwardneural
neuralnetwork
networkprediction
predictionmodel
modelconstructed
constructedbased
based
onthe
on themulti-objective
multi-objective dataset
dataset of of rural
rural residential
residential buildings
buildings in cold
in cold regions,
regions, the system
the system code
code was encapsulated to enhance its usefulness for designers. The software was built
was encapsulated to enhance its usefulness for designers. The software was built with
with Tkinter,
Tkinter, a standard
a standard GUI for
GUI library library for Python,
Python, and mainlyandpredicted
mainly predicted the three
the three target target
variables.
The interface of the software is shown in Figure 15.
variables. The interface of the software is shown in Figure 15.

Figure15.
Figure 15.Software
Softwareuser
userinterface.
interface.

4.4.Results
Resultsand
andDiscussion
Discussion
After the model training is completed, we save the trained model. By plotting the
comparison line graphs of the actual values versus the predicted values for the three
types of prediction targets in the test and validation sets(Figure 16), the predictive model
demonstrates a relatively good level of prediction for all three targets. Notably, the model’s
performance in predicting COST is better than the other two indicators.
Then, to test the saved model’s accuracy, the saved regression prediction model is
imported, and three groups of variables are randomly selected as cases in each type of RRB
for prediction. The differences between the prediction results and the actual performance
levels are compared. The case parameter variables are shown in Table 4. A comparison of
actual and predicted target values is shown in Table 5.
As shown in Table 5, the model shows a good prediction performance for both TLI
and UTCI, and the prediction for COST sometimes produces large fluctuations. However,
overall, the model predictions are still accurate, and the prediction accuracy should be
higher if more samples are provided in the multi-objective optimization stage, and more
case data are generated.

Table 5. Random case target value.

Real Value Predicted Value Variation Percentages

TLI COST UTCI TLI COST UTCI TLI COST UTCI
1 99.588 15,975.544 17.2356 100.945 17,254.183 17.232 0.013 0.080 0.0002
2 94.141 15,672.962 17.2003 93.556 14,721.819 17.175 0.006 0.061 0.0014
3 73.972 18,869.712 17.0575 76.573 18,588.224 17.017 0.035 0.014 0.0023
4 107.564 34,047.649 16.6412 103.649 35,139.878 16.603 0.036 0.032 0.0022
5 96.054 19,117.673 16.4816 96.596 19,841.098 16.504 0.005 0.037 0.0013
6 104.105 17,278.576 17.0768 105.630 17,330.527 17.083 0.014 0.003 0.0003
7 118.379 26,361.811 16.0661 121.534 27,027.687 16.042 0.026 0.025 0.0015
8 82.548 29,951.88 16.7105 80.286 30,869.421 16.689 0.027 0.031 0.0012
9 76.796 40,781.825 16.4246 79.623 42,006.448 16.430 0.036 0.030 0.0003

Sustainability 2024, 16, 667
After the model training is completed, we save the trained model. By plo ing the
comparison line graphs of the actual values versus the predicted values for the three types
of prediction targets in the test and validation sets(Figure 16), the predictive model
demonstrates a relatively good level of prediction for all three targets. Notably, the
16 of 19
model’s performance in predicting COST is be er than the other two indicators.

Figure
Figure 16.
16. Comparison
Comparison of
of the
the true
true value
value of
of the
the forecasted target with
forecasted target with the
the forecasted
forecasted value.
value.

Then, to test the

In conclusion, thesaved model’s
machine accuracy,
learning modelthe hassaved regression
high accuracy in prediction model
predicting and is
eval-
imported,
uating the and three groups
economic of variables
and energy areperformance
efficiency randomly selected as cases
of courtyard in each type
buildings, of
among
RRB forCompared
others. prediction.with
The differences between
the traditional the prediction
performance resultswhich
simulation, and the actual
takes 3–5 perfor-
min to
mance
simulate levels
threeare compared.
targets The
at a time case parameter
(depending on thevariables
performanceare shown in Table 4.it A
of the computer, com-
may be
parison
reducedof oractual and predicted
increased), targetmodel
the prediction valuescan
is shown
obtain in
theTable 5.
prediction results immediately
by inputting the parameters, which significantly improves the efficiency at the early stage
Table 5. Random
of design. case target
However, when value.
specific datasets are not large enough, the prediction accuracy is
relatively low, suggesting the need for further improvements, such as better model tuning
Real Value Predicted Value Variation Percentages
or the enhancement of the data collection and processing process.
TLI COST UTCI TLI COST UTCI TLI COST UTCI
1 99.588 15,975.5445. Conclusions
17.2356 100.945 17,254.183 17.232 0.013 0.080 0.0002
In this study, a neural network model that can accurately predict the energy efficiency,
as well as the economy of rural residential buildings, in the cold regions of China, was de-
veloped by focusing on the energy consumption and envelope spending of rural residential
buildings in China, using examples from the rural Jinan area. The model is developed based
Sustainability 2024, 16, 667 17 of 19

on multi-objective optimization and machine learning and broadly applies to equivalent

building types.
The research methodology can be applied in rural planning as well as building renewal.
Given the unbalanced nature of rural construction, the machine learning model in this
method can help designers quickly assess the performance of buildings at the early stage
of updating and evaluating buildings, avoiding irrational renovations and additions and
maximizing low-carbon updating. Research and data collection of existing residential
buildings in Dongquan Village established a performance simulation parameter model.
Then, using a genetic algorithm for multi-objective optimization, a total of 6000 datasets
containing spatial parameters and target performance were generated for three residential
buildings. The neural network model was then constructed and trained to evaluate and
predict the changes in energy consumption, cost, and thermal comfort of the courtyard due
to changes in building and courtyard parameters. According to the model validation results,
R2 = 0.988 and MSE = 0.0148, indicating that the model can correctly predict the target
variables of the design solution. The code was also encapsulated into a more user-friendly
interface presentation for designers.
This study’s innovation is combining the neural Network with MOO and present-
ing it in a designer-friendly interface that shortens the simulation time and improves the
evaluation efficiency compared with the traditional method using building performance
simulation software. The model allows the fast prediction of the same type of building
under the same area, and the method is also generalizable. Theoretically, the method can be
applied to buildings of ensemble type in various regions, but modeling based on different
meteorological data in different regions is still needed for accurate prediction. In addition,
in the next step of the study, other indicators and variables can be incorporated into the eval-
uation system, and through a complete prediction and evaluation methodology, it is hoped
that it will provide references and ideas for the future renewal of the vast number of rural
residential buildings, which will help achieve the broader goal of sustainable development.

Author Contributions: Conceptualization, Y.X. and M.S.; Methodology, Y.X. and M.S.; Software, M.S.;
Validation, M.S.; Formal analysis, M.S.; Investigation, M.S. and L.W.; Resources, M.S.; Writing—original
draft, M.S.; Writing—review and editing, M.S.; Visualization, M.S.; Supervision, Y.X.; Project administra-
tion, Y.X. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: At this time, the data and source code for the machine learning models
produced in this study are not available as they are required for the subsequent phase of the research.
Conflicts of Interest: The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this paper.

Nomenclature
RRB rural residential buildings EC energy consumption
IPCC intergovernmental panel on climate change CES cost of the enclosure structure
MOO multi-objective optimization RF roof slope
PHW principal house width BO building orientation
PHD principal house depth HCW height of the courtyard’s wall
WHW wing house width EPS EPS thickness
WHD wing house depth XPS XPS thickness
BH building height WD window type
EW eave width TLI total loads intensity
ML machine learning UTCI Universal Thermal Climate Index
WIC width of the courtyard LIC “length of the courtyard”
Sustainability 2024, 16, 667 18 of 19

References
1. Merlet, Y.; Rouchier, S.; Jay, A.; Cellier, N.; Woloszyn, M. Integration of phasing on multi-objective optimization of building stock
energy retrofit. Energy Build. 2022, 257, 111776. [CrossRef]
2. China Building Energy Efficiency Association Building Energy Consumption and Carbon Emission Data Specialized Committee.
Efficiency, Research Report on Building Energy Consumption in China; Chinese Construction Industry Chubns: Beijing, China, 2022.
Available online: https://www.cabee.org/ (accessed on 23 June 2023).
3. Building Energy Efficiency Research Center, Tsinghua University. China Building Energy Efficiency Development Research Report
2023; China Building Industry Press: Beijing, China, 2023.
4. Wu, Y. Research on Low Carbon Adaptive Construction Methods and Practices of Rural Community Spatial Form. Ph.D. Thesis,
Zhejiang University, Zhejiang, China, 2016.
5. He, H.; Xiong, K.; Li, Y. Status and outlook of the quality of concrete residential projects in villages and towns in China. Concrete
2010, 247, 129–131.
6. Liang, L.; Wu, W.L.; Lal, R.; Guo, Y.B. Structural change and carbon emission of rural household energy consumption in Huantai,
northern China. Renew. Sustain. Energy Rev. 2013, 28, 767–776. [CrossRef]
7. Wang, R.; Jiang, Z. Energy consumption in China’s rural areas: A study based on the village energy survey. Clean. Prod. 2017, 143,
452–461. [CrossRef]
8. Fan, J.L.; Yu, H.; Wei, Y.M. Residential energy-related carbon emissions in urban and rural China during 1996–2012: From the
perspective of five end-use activities. Energy Build. 2015, 96, 201–209. [CrossRef]
9. Xu, J.; Gao, W.J.; Huo, X. Analysis on energy consumption of rural building based on survey in northern China. Energy Sustain.
Dev. 2018, 47, 34–38.
10. He, B.-J.; Yang, L.; Ye, M. Buildings energy efficiency in China rural areas: Situation, drawbacks, challenges, corresponding
measures and policies. Sustain. Cities Soc. 2014, 11, 7–15. [CrossRef]
11. Jermyn, D.; Richman, R. A process for developing deep energy retrofit strategies for single-family housing typologies: Three
Toronto case studies. Energy Build. 2016, 116, 522–534. [CrossRef]
12. Zhang, Y.; Yang, L. Comparative study on the indoor thermal environment of traditional and modern residential houses in
northern Jiangsu Province. Archit. Technol. 2017, 48, 707–710.
13. Asadi, E.; Da Silva, M.G.; Antunes, C.H.; Dias, L. Multi-objective optimization for building retrofit strategies: A model and an
application. Energy Build. 2012, 44, 81–87. [CrossRef]
14. Liu, Y.; Huang, Z.; Wang, Y. Energy-efficient design of a farmhouse based on seasonal differences in indoor and outdoor
multi-space thermal environment quality and user space utilization. Build. Sci. 2022, 38, 216–222.
15. Al-Masri, N.; Abu-Hijleh, B. Courtyard housing in midrise buildings: An environmental assessment in hot-arid climate. Renew.
Sustain. Energy Rev. 2012, 16, 1892–1898. [CrossRef]
16. Taleghani, M. Outdoor thermal comfort by different heat mitigation strategies—A review. Renew. Sustain. Energy Rev. 2018, 81,
2011–2018. [CrossRef]
17. Zamani, Z.; Heidari, S.; Hanachi, P. Reviewing the thermal and microclimatic function of courtyards. Renew. Sustain. Energy Rev.
2018, 93, 580–595. [CrossRef]
18. Tavakolan, M.; Mostafazadeh, F.; Eirdmousa, S.J.; Safari, A.; Mirzaei, K. Parallel computing simulation-based multi-objective
optimization framework for economic analysis of building energy retrofit: A case study in Iran. J. Build. Eng. 2022, 45, 103485.
[CrossRef]
19. Gao, Y.; Hu, K.; Yue, X.; Yuan, J. Multi—Objective optimization design of form parameters for energy saving and thermal comfort
in northern country houses. Overseas Chin. Univ. Sci. (Nat. Sci. Ed.) 2021, 42, 619–627.
20. Vukadinović, A.; Radosavljević, J.; Ðord̄ević, A.; Protić, M.; Petrović, N. Multi—Objective optimization of energy performance for
a detached residential building with a sunspace using the NSGA-II genetic algorithm. Sol. Energy 2021, 224, 1426–1444. [CrossRef]
21. Ascione, F.; Bianco, N.; De Masi, R.F.; Mauro, G.M.; Vanoli, G.P. Design of the building envelope: A novel multi-objective
approach for the optimization of energy performance and thermal comfort. Sustainability 2015, 7, 10809–10836. [CrossRef]
22. Carlucci, S.; Cattarin, G.; Causone, F.; Pagliano, L. Multi—Objective optimization of a nearly zero-energy building based on
thermal and visual discomfort minimization using a non-dominated sorting genetic algorithm (NSGA-II). Energy Build. 2015, 104,
378–394. [CrossRef]
23. Yan, H.; Yan, K.; Ji, G. Optimization and prediction in the early design stage of office buildings using genetic and XG Boost
algorithms. Build. Environ. 2022, 218, 109081. [CrossRef]
24. Zekić-Sušac, M.; Has, A.; Knežević, M. Predicting energy cost of public buildings by artificial neural networks, CART, and
random forest. Neurocomputing 2021, 439, 223–233. [CrossRef]
25. Guo, J.; Zhou, J.; Li, M.; Lu, S. Based on ANN and many-objective optimization to improve the performance and economy of
village houses in Chinese cold regions. J. Build. Perform. Simul. 2023, 16, 526–536. [CrossRef]
26. Liu, C.L.; Tseng, C.J.; Huang, T.H.; Yang, J.S.; Huang, K.B. A multi-task learning model for building electrical load prediction.
Energy Build. 2023, 278, 112601. [CrossRef]
27. Lagios, K.; Niemasz, J.; Reinhart, C.F. Animated building performance simulation (ABPS)–linking Rhinoceros/Grasshopper with
Radiance/Daysim. Proc. SimBuild 2010, 4, 321–327.
Sustainability 2024, 16, 667 19 of 19

28. Tian, Y.; Huang, Q. Multi-objective optimization design method for building performance and its application—An example of
genetic algorithm. New Archit. 2021, 198, 84–89.
29. Jiang, J. Research on the Optimization of High-Density Neighborhood Morphology Design Based on Genetic Algorithm. Master’s
Thesis, Shenzhen University, Shenzhen, China, 2020.
30. Nelli, F. Machine Learning with sci-kit-learn. In Python Data Analytics; Apress: Berkeley, CA, USA, 2018.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.