Improved Genetic Algorithm Used In Elevator Group Control Systems

Elevator, as vertical transporting facility is more and more widely used with the emergence of intelligent buildings. The quality of elevator services needs to be improved to satisfy the passengers'requirements. Single elevator is insufficient for the inner building transportation. Multi-elevator service could shorten the waiting time and reduce energy consuming. As a consequence, the elevator group control system is introduced. The elevator group control system improves the service efficiency and quality by an optimized control strategy, which harmonizes multi-elevator operation.The paper reviewed the development and current status of EGCS in domestic and international, and introduced advantages and shortcomings of a variety of EGCS dispatch method. We introduced characters of the elevator group control system, such as many-targets, the uncertainty, and the non-linear characteristic. The paper analyzes the necessity and the feasibility of applying the genetic algorithms to the elevator group control system. This paper optimize the control objectives such as shorter waiting time and riding time, less crowd, less power consumption. The evaluation function is constructed by the weighted combination of several optimization objectives based on combinatorial optimization method. The multi-objective dispatching of elevators is realized by searching the evaluation function for the best dispatching scheme with the genetic algorithm. In order to improve the search efficiency of Genetic Algorithm, an optimum maintaining strategy is applied to the operating process of Genetic Algorithm. An improved adaptive genetic algorithm, in which the probability of crossover and mutation were varied depending on the fitness values of the solutions, is applied to increase the performance of convergence to the global optimum.The elevator group control pessanger traffic flow is established by analyzed the transport process of EGCS. The EGCS simulation model is established in the text with the C++ language. The simulation results show that various targets meet the requirements according to various simulation data.