Republic of the Philippines

Polytechnic University of the Philippines

College of Computer and Information Sciences
Department of Computer Science

“Artificial Neural Networks based Traffic Light Controller”

In Partial Fulfillment for the subject
CS ELEC 2: Computational Intelligence

Submitted By:
Sarcia, Bella S.
Fulo, Alyssa Marie
Barrameda. John Carl
Pelaez, Zaldrich A.
Montero, Rexie N.
Perez, Patricia Lj
Lascano, Billy Geoffertson A.
Costa, Samuel

Bachelor of Science in Computer Science 3– FS2

March 30,2019

Submitted to:
Prof. Ranil Montaril
I.Abstract

(After the completion of all of these)

II. Background of the Study

A. Key Terms

 Artificial Neural Network-is a computational model based on the structure and functions

of biological neural networks. Information that flows through the network affects the

structure of the ANN because a neural network changes - or learns, in a sense - based on

that input and output.

 Sigmoid Function-  is a bounded, differentiable, real function that is defined for all real

input values and has a non-negative derivative at each point.

 Tansig Function-  Hyperbolic tangent sigmoid transfer function.

 Feed Forward Algorithm- also known as “Feed Neural Network”. It was the first and

simplest type of artificial neural network devised[2]. In this network, the information

moves in only one direction, forward, from the input nodes, through the hidden nodes (if

any) and to the output nodes. There are no cycles or loops in the network.
B. Problem Statement

To create an intelligent agent that will control the Traffic Light State

transition and timing control. For the ANN to suggest the next Traffic light state

and timings on the two traffic lights on each road direction/intersection based on

the current Traffic light state.

C. Applicable Related Studies

According to the research of [Choy et al.] Neural Networks for Real-Time Traffic Signal

Control, Real-time traffic signal control is an integral part of the urban traffic control system, and

providing effective real-time traffic signal control for a large complex traffic network is an

extremely challenging distributed control problem, the researchers used Neural Network for their

presented work. The SPSA-NN model was applied to control traffic signals for a large traffic

network with multiple intersections, using a three-layer NN and taking relevant traffic variables

as inputs.

ANN can be viewed as weighted directed graphs in which artificial neurons are nodes

and directed edges (with weights) are connections between neuron outputs and neuron inputs.

Based on the connection pattern (architecture), ANN can be grouped into two categories

* feed-forward networks, in which graphs have no loops and

* recurrent (or feedback) networks, in which loops , and occur because of feedback

connections.
 In the most common family of feed-forward networks, called multilayer perceptron,

neurons are organized into layers that have unidirectional connections between them. Figure 1

also shows typical networks for each category. Different connectivities yield different network

behaviors. Generally speaking, feed-forward networks are static, that is, they produce only one

set of output values rather than a sequence of values from a given input. Feedforward networks

are memory-less in the sense that their response to an input is independent of the previous

network state. Recurrent, or feedback, networks, on the other hand, are dynamic systems. When

a new input pattern is presented, the neuron outputs are computed. Because of the feedback

paths, the inputs to each neuron are then modified, which leads the network to enter a new state.

Different network architectures require appropriate learning algorithms. The next section

provides an overview of learning processes [Jain et al.]

Figure 1. A taxonomy of feed-forward and recurrent network architectures.

A feedforward neural network is an artificial neural network wherein connections

between the nodes do not form a cycle.  Zell, Andreas (1994) As such, it is different

from recurrent neural networks.

 The feedforward neural network was the first and simplest type of artificial neural

network devised. In this network, the information moves in only one direction, forward, from the

input nodes, through the hidden nodes (if any) and to the output nodes. There are no cycles or

loops in the network.  Zell, Andreas (1994).

III. Software Design Architecture

A .Applicable Equation

(Rexie)
B. Functional Block

Figure 1. Functional Block Diagram

Input data set contains the case for the traffic lights like time and traffic light color it will then

undergo learning process using Feed Forward algorithm. Neural Network has 2 inputs, 1

hidden layer which consist of 3 neurons. This hidden layer is wherein the sigmoid function is use
as a transfer function to the output layer. The output layer has the tansig function as a transfer

function to the "Output". As soon as the condition is given the Traffic Light interface

immediately display the output.

C. Sequence Diagram/State Diagram

(Rexie)

D. ANN Architecture

The Neural Network will be a Feedforward neural network having two inputs and one

output.

 The input layer comprise of two neurons X1, X2.

 The hidden layer comprise of three neurons H1, H2, H3, with activation function

Sigmoid.

 The output layer comprise of one neuron Y1, with activation function Tansig.
HIDDEN
LAYER

H1
INPUT OUTPUT
LAYER LAYER

INPUT X1
1

H2 Y1 OUTPUT
1

INPUT X2
2

H3 Activation Function:
Tansig

Activation Function:
Sigmoid
IV. Simulation/Test Result

V .Conclusion

(After Test Result)

VI. Recommendation

The researchers believe that the system could further be enhanced through the use of

other algorithms that could consider the problems of traffic volume in coordination with the

traffic light controller. That is why; the researchers recommend that the future researchers could

use the combination of Genetic Algorithm in conjunction with the approach of Artificial Neural

Networks with traffic lights controllers where the Genetic Algorithm could be utilized in finding

the most optimal time transition given a certain volume of vehicles. So that the traffic light

controller would be more beneficial in the enhancement of traffic congestions.

VII. References
Zell, Andreas (1994). Simulation Neuronaler Netze [Simulation of Neural Networks] (in

German) (1st ed.).

Anil K. Jain, Jianchang Mao, KM. Mohiuddin, “Artificial Neural Networks: A Tutorial” (1996).

Dipti Srinivasan, Min Chee Choy, and Ruey Long Cheu, “Neural Networks for Real-Time

Traffic Signal Control” (2006)