17-04-2013, 03:16 PM
AN INTELLIGENT TRAFFIC LIGHT ALGORITHM SIMULATION
AN INTELLIGENT TRAFFIC.doc (Size: 246 KB / Downloads: 47)
Background of Traffic Management in Nairobi
The idea of traffic light controller using fuzzy logic is an adaptation from conventional traffic light control system. Traffic light is an important system to control the traffic flow especially at junctions. But, we find some problem with the conventional traffic light control systems at times. Conventional traffic light cannot operate as efficiently as expected. Because of that problem, the idea to develop an intelligent traffic light control system based on fuzzy logic is taken into consideration (Ahmad, 2005).
The essence of any urban transportation problem is lack of mobility, severely limited mobility and mobility purchased at a very high social and economic cost. In Nairobi, the current situation of urban transport is alarming. Despite the relatively low levels of private automobile ownership, the city’s transportation problems are severe in degree, daily duration and areas affected .These problems are especially felt during the peak demand hours which are often characterized by considerable jostling and stampeding among the travelling public in search of the means of public transport. The chaotic situation is further exacerbated by the carelessness and apparent lack of concern among the public service vehicle (PSV) operators (Obudho, 1993).
Problem Statement
Right now, the length of time it takes to move from one part of the city of Nairobi to another has reached a totally inefficient point. There is hardly any difference between the volume of traffic on the road at peak hours and off-peak hours. We can conclude this is a traffic management problem because sometimes, the roads are clear during off-peak hours as you expect them to be, and even during peak hours, yet there is no plausible reason to explain this anomaly (Butoy,2009).
The existing environment consists of traffic lights, which are conventional. The traffic lights, in the absence of a vehicle on any particular session, will continue to operate as if traffic always exists and assume an equal distribution of traffic flow.
Research Justification
There have been increased cases of congestions due to poor traffic controls in our Roads particularly at our roundabout junctions. This has necessitated the need for effective and efficient traffic control systems to alleviate the fears of the Government and uncertainty of the travel times by motorists. The research will address the congestion problem in our Roads. With the proposed solutions, motorists and pedestrians will be able to spend less time travelling and will not be frustrated or delayed when travelling.
Pre-timed Signal Control Algorithm
The control (signal plan) is calculated in advance, using statistical data (Askerzade, 2006). The control uses preset cycle time to change the lights (fixed time). The general structure of the preset traffic control is shown in fig.2.1. The main control measure in urban road networks is the traffic lights at intersections. Traffic lights besides ensuring the safety of road crossings may also help in the minimization of total time spent by all vehicles in the network, provided that an optimal control strategy is applied.
Actuated Signal Control Algorithm
Askerzade (2006) stated that the real-time data about traffic processes are used to determine control or its modification. The control combines preset cycle time with proximity sensors which can activate a change in the cycle time or the lights. Actuated signal control, one of the most widely deployed traffic control strategies, taking advantage of the data collected by the detectors, is more adaptable to the real traffic condition. The signal control decision is made according to a set of rules considering the traffic condition.
Neural Network Algorithm
The human brain, according to Bradley (2004) is constructed of cells called neurons. Each cell accepts some inputs and then, based on the total value of inputs, the neural decides whether or not to fire an output. In the brain vast numbers of neurons are wired together, sending their outputs to other neurons, and ultimately allowing humans to make complex decisions about things. Neural networks attempts to replicate this process electronically.
Neural network follow the same architecture as the brain, except the neurons are represented electronically. In a biological neuron, messages pass from cell to cell over gaps called synapses. Input messages then travel along dendrites. After the cell generates its output, an output message is sent out along an axon. Just as with biological neurons, neuron in a neural network has a set of inputs that it accepts then uses to calculate its outputs.
Genetic Algorithm
Ayad (2009) introduces Genetic algorithm in the traffic light control system to provide an intelligent green interval response based on dynamic traffic load inputs, thereby overcoming the inefficiencies of conventional traffic controllers. In this way the challenges are resolved as the number of vehicles are read from sensors put at every lane in a four-way, two lane junction and pedestrians are motivated at the road junction. The features inherent in genetic algorithm play a critical role in making them the best choice for practical applications, namely optimization, computer aided design, scheduling, economics and game theory. It is also selected because it does not require the presence of supervisor or observer. However, genetic algorithm without prior training, continuously allow permanent renewal of decisions in generating solutions. Instead of trying to optimize a single solution, they work with a population of candidate solutions that are encoded as chromosomes. These chromosomes are separate genes that represent the independent variables for the problem at hand.
Fuzzy Rule-Based Traffic Flow Algorithm
Syed (2009) stated that conventional methods of traffic signal control based on precise models fail to deal efficiently with the complex and varying traffic situation. They are modeled based on the preset cycle time to change the signal without any analysis of traffic situation. Due to fixed cycle time, such systems do not consider which intersection has more load of traffic, so should be kept green much longer or should terminate earlier then complete cycle time. Fuzzy rule-based controllers are proved to be well managers of traffic light system in such scenarios. Fuzzy controllers have the ability to take decision even with incomplete information. These algorithms are continually improving the safety and efficiency by reducing the waiting times of vehicles. These increases the tempo of travel and this makes signals more effective and traffic flow smooth.