Application of evolutionary algorithms in traffic light scheduling

• Main Author: Tang, Shi Jie
• Other Authors: Su Rong
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/69298
• Institution: Nanyang Technological University

This report address urban traffic light scheduling problem (UTLSP). The urban traffic light control problem is describe by a centralized model in a scheduling framework. The concept of splits, offsets and cycles are not taken consideration in the proposed model, therefore the UTLSP is under the optimization problems. The network controller assigned each traffic light in a real-time manner. This project aim to minimize the network-wise total delay time in a given finite horizon. To solve this problem, the Jaya algorithm was proposed. An improvement strategy is also proposed to better the results. A feature based search operator is utilized to improve the search performance of the optimization method. The experiments are carried out based on the real traffic data in Singapore. The Jaya and iJaya algorithms are evaluated by solving twenty-seven cases of large-scale traffic network. The comparisons and discussions proven that the evolutionary algorithms can effectively use to solve the UTLSP.