2018 Volume 9 Issue 1 Pages 49-59
In recent years, many researchers have become interested in methods for mitigating traffic congestion by optimizing traffic signal parameters. To mitigate traffic congestion over a widespread area, a method using an advanced genetic algorithm and a traffic simulator has been proposed (Nishihara, T., et al., “The Verification with Real-World Road Network on Optimization of Traffic Signal Parameters using Multi-Element Genetic Algorithms”, ITS World Congress, 2012). However, this method consumes considerable time when simulating traffic flow. This paper proposes a method that reduces the processing time of the simulator by using a neural network.