Abstract:
For elevator group control system (EGCS) with a complicated optimization and decision problems, a reinforcement learning(RL) model for EGCS is built using multi-agent theory and technology in this paper. Furthermore, an algorithm for RL based on the genetic algorithm (GA) is proposed and the general descriptive algorithm is also given. The virtual simulation environment for EGCS is established. The simulation results show that the proposed GA-RL algorithm is valid for promoting the efficiency and the convergence speed of the RL algorithm and improving the population structure.