Research On Intelligent Dispatch Of Elevator Group Control System Based On Neural Network

The elevator is an indispensable vertical traffic means of transport in the skyscraper, with the emerging in a large amount of the high-story intelligent building, passengers have put forward higher and higher demand to the systematic function of the elevator. In recent years, the control technology of the elevator has got fast development, a lot of advanced control technology is applied to the Elevator Group Control System (EGCS), improving the systematic performance of EGCS greatly, but there are many stressing problems demanding prompt solutions.The thesis regards the EGCS as the research object. EGCS is a complex, nonlinear, uncertain, multi-objective and random decision-making system, this paper has introduced the system's control fashion and performance evaluate guidelines.Then this paper has introduced the up-peak traffic condition, down-peak traffic condition, multi-way traffic condition, balanced floor traffic condition and free traffic condition. It also proposes control strategies to the different traffic conditions. It can improve the system's performance.The article has deeply researched the dispatch algorithm of the EGCS based on the detail analyzed in the passenger flow characteristic description of transportation variable. Simultaneously this article take enhancing the elevator group control system the intellectualized level as a starting point, by passenger's average time of waiting for the elevator short, the passenger long time waiting rate is low, as well as the system movement energy consumption few three important targets average weighted to structure the evaluating function, and adjustment corresponding weighting coefficient according to different transportation patterns . Compared to the traditional evaluating function, this function can evaluate the elevator's performance more accurate and completely. The elevator group control system use the self-study of the neural network to dispatch the elevator, choose different dispatch algorithm according to the traffic flow change of the. building in different time. The neural network can adjusts the weights value through self-study, causes the control strategy of the entire EGCS to be at the optimistic condition.Finally this article has constructed the simulation model of the building elevator group control system, use MATLAB to simulate the elevator group control algorithm, the result of simulation had indicated that multi- goals intelligence dispatch method based on the neural network used in the EGCS could obtain the reasonable dispatch result, it can enhance the operating efficiency of the system and increases the passenger's satisfaction degree.