الملخص الإنجليزي
Over the years, Genetic Algorithm (GA) has become one of the powerful methods to solve different optimization problems such as Travelling Salesman Problem (TSP), the Vehicle Routing Problem (VRP), the Sudoku Problem, and the Signal Timing Problem.
In recent years, the optimization of traffic signal timing has been one of the main concerns of traffic engineers. Finding an intelligent solution will reduce not only the overall traffic, but also vehicle delays, vehicle emissions and fuel consumption at an intersection. Different algorithms, including heuristic algorithms such as Hill Climbing Algorithm (HCA), Simulated Annealing Algorithm (SAA), Ant Colony Algorithm (ACA) etc., are used to develop traffic light control systems (TLCS).
GA has been used in several works to control traffic light systems. It approved its strength in optimizing timing plans for an intersection. This thesis presents an improved GA to optimize traffic signal timing. It introduces a new crossover operator that attempts to enhance the GA performance and locates the optimal or near optimal solutions.
To verify the success of the proposed algorithm a standard genetic algorithm was used to confirm the merits of the new crossover operator. The results reveal that the GA that uses the new crossover operator can generate better timing plans than the timing plans generated by the standard GA that uses one point crossover operator for more than 40% of the tested instances. Moreover, the new traffic control system can manage the flow of vehicles and reduce the overall traffic in an intersection efficiently compared to the fixed time TLCSs.
