A Study on Metaheuristics for Urban Traffic Light Scheduling Problems

Abstract

Metaheuristics is the method used to complete complex optimizations. Metaheuristics is also used in urban traffic lights. This paper will discuss the metaheuristics used in urban traffic lights and divide it into two approaches. First, improve the performance of metaheuristic algorithms by coordinating between traffic intersections. Second, change the fixed time signal controller into an adaptive time signal control, then use metaheuristics to get optimal green signal time. The first approach is done by improving the performance of the algorithm by finding the smallest ARPD value of the improved Artificial Bee Colony (ABC) Algorithm and Harmony Search (HS). The result obtained the smallest average ARPD to 0.74 for the Artificial Bee Colony Algorithm and the computing time for the largest case (10 * 10 for 60s) using Discrete Harmony Search (DHS) was 6,219s. While completion of the second approach managed to raise the vehicle speed up to 27.6%. In addition, this paper insights the performance comparison of metaheuristic algorithms to solve the problem of scheduling urban traffic lights.