Traffic Signal Optimization with Estimation of Distribution Algorithms and VISSIM-MATLAB Integrated Simulation Platform

Surrogate-assisted cooperative signal optimization for large-scale traffic networks

“…Experimental results showed that the employed algorithm indicated better performance than the simulated annealing algorithm in reducing the travel time of vehicles. Liang et al [33] introduced two modified estimation of distribution algorithms for traffic signal optimization to reduce the total vehicle delay time. Experiments on a traffic network with 11 intersections indicated that the two modified EDAs could outperform several existing mate-heuristic algorithms.…”

“…Existing research works mainly focus on optimizing signals for crossroads [3,4,33] Considering that the purpose of signal optimization is to find suitable signal cycles for all junctions and each signal cycle is composed of several signal phases, we encode a signal plan into a vector of integers, where each integer represents a signal phase duration time in a specified signal cycle and is limited to a certain range, ensuring that vehicles in different roads can safely cross a junction within reasonable time. To make it more clear, a simple solution encoding example is presented in Table 1.…”

“…Although the above experimental environment cannot perfectly simulate real traffic situations, it is still very meaningful to verify the efficiency of SCSO. Similar experimental setup method was also widely used in many existing works [7,33,40].…”

“…To verify the efficiency of SCSO, we compare it with six mate-heuristic algorithms for fixed-time signal optimization, including ABC, DHS and Jaya [3], PSO [7], TSO [4] and mEDAve [33]. These algorithms have all been briefly introduced in Section 2.1.…”