Study On Elevator Group Control Based On Fuzzy Neural Network

This paper is the study of the intelligent algorithm of elevator group control system. The research direction is the use of fuzzy neural network to realize the intelligent elevator group control system dispatching, to save time and energy consumption, improve the service quality.This paper outlines the elevator development history and research situation at home and abroad, analysis lists some commonly used intelligent elevator group control system algorithms, such as fuzzy logic, artificial neural network, expert system, genetic algorithm and so on. Through analyzing and comparing the advantages and disadvantages of each elevator group control algorithm, finally we choose the method of fuzzy neural network. The control algorithm defines current traffic mode of the building according to traffic feature value of the elevators. Different weights is set basing to different traffic modes. The fuzzy neural network will calculate the trusting values of each elevator to response to certain call, then make the dispatching according to the result.By the long term statistics of the traffic flow in the high buildings, and the further investigation of the statistics, this paper summarized the traffic features of the buildings. And this paper defines the traffic mode of elevators into up-going peak traffic mode, down-going peak traffic mode, idle traffic mode, inter floor traffic mode, two-way traffic mode and four way traffic mode, according to the traffic features. In this paper, fuzzy neural network is used to define the traffic mode of the elevator group. Then the fuzzy neural network is used again in the dispatching algorism to calculate the trusting value of responses from each elevator. The most trustable elevator is selected to response the call and make the ride. This paper used MATLAB software for simulation, and a good result is gained.