| Summary: | Traffic signal optimization for both traffic plan design and signal timing is a task that demands significant time and computational resources. In this paper, we formulate a rapid algorithm for finding the best combination of signaling phases and their optimal timings which minimize the total delay of the entire period based on the multiple-period analysis proposed in Highway Capacity Manual 2010. The optimization problem is solved in MATLAB using two optimization algorithms: the genetic and the scatter search algorithms. Comparative evaluation of the algorithms in relation to their estimation time and their solutions was performed at a reference intersection. The estimated time-dependent delay was calibrated and validated in AIMSUN microsimulation software for the base conditions and the optimized plans. The comparison of the global optimal solution of the two algorithms and the true global optimal was performed through a computational experiment in a two approach intersection with proper discretization of the variables. The results showed that the scatter search algorithm is superior to the genetic algorithm with respect to the required runtime and the estimated optimal solution. Keywords: Traffic control optimization, Global optimization, Genetic algorithms, Scatter search algorithms, Microscopic simulation
|
|---|
