Design and development of neuro-fuzzy techniques for monitoring traffic conditions

• Main Author: Meng, Detong
• Other Authors: Er Meng Joo
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/68817
• Institution: Nanyang Technological University

In modern society nowadays, the transportation is playing an increasingly crucial role both in economic development and urbanization. From Singapore’s perspective, the transport system in fact is effective and highly developed. However, in recent years with the rapidly increasing of vehicles, the public transportation is now critically confronting the issues such as overcrowding and congestion. The tradition method used to handle the traffic conditions is to manage the traffic light through setting up the pre-timed controller. But obviously, these traffic infrastructures cannot stratify the requirement any more. In order to solve this issue, a new effective technology called neural fuzzy network is introduced in our project. In real time traffic conditions, this method can be widely and effectively used in the intersection traffic signal control. In addition, compared with the fuzzy expert system and artificial neural network, this method is more advanced and friendly-used. The main idea for this method applying in traffic field is achieved by controlling the traffic lights to adjust to the optimum state which can satisfy various changing conditions including congestion, oversaturated and accident etc. Therefore, by introducing the neural fuzzy network, we will try to promote the traffic conditions by improving the throughput of the vehicles though extending the green light period when necessary and minimizing the delay as well. More precisely, in our project, a technique called Generalized Dynamic Fuzzy Neural Network is being used in a purpose to obtain the fuzzy rules which will be used in the traffic controller. Therefore, we also pay lots of attention to the controller utilizing the fuzzy logic control. By studying fuzzy logic controller, we will try to develop a link between these two fields. Due to limited conditions, we mostly study this project by using the MATLAB, and a simulated traffic intersection is design. In addition, a series of simulated traffic data is generated for studying and training purpose.