Simulation-Based Heuristic Optimization Of A Traffic Systems

This paper deals with nonlinear optimization of a stochastic system via simulation and a heuristic algorithm. These tools are used for optimization of the time parameters of the traffic lights of three junctions at Konečného square in Brno, Czech Republic, in order to reach maximum possible throughput. The objective is to minimize average waiting time in the system that might be described as an open queuing network. This is done in two steps: building a simulation model of Konečného square in Java using SSJ (Stochastic Simulation in Java a Java library for stochastic simulation) and implementing a heuristic algorithm Simulated Annealing that is using the simulation model for optimization. After a brief description of the traffic system, its mathematical and simulation models are introduced. Then the way of getting input data is discussed as well as verification and validation of the simulation model. The results of an optimization based on Simulated Annealing are shown and interpreted in the last part of the paper. Another contribution of building the simulation model is evaluating the effect of implementing the preference given to public transportation.