| With China’s sustained economic steady and fast development, the elevator has become an indispensable supporting facility, timely detection of faults and timely maintence failure is the key to improve the level and ensure personal security.The elevator control system is the most important part in the ac variable frequency variable voltage speed control elevator system, which is part of the elevator fault often happens.Currently, elevator fault diagnosis method is mainly rely on feeling and experience of the technical personnel, there is no guarantee that the elevator fault diagnosis of rapidity and precision. In order to solve this situation, this paper expend the research of elevate control system’s fault diagnosis.Purpose is that when the elevator down to diagnose the fault in time, ensure the safety of personnel and equipment. The main contents are as follows:First, considering the entire elevator system and its complexity, and the status of acquisition are very large, so research on elevator control system in this article. Fault signal of the elevator control system is uncertainty, fuzziness, and non-linearity. At the same time, taking into account the data collection having a periodicity as a result of cannot be used for real-time signal acquisition, in this paper, combine fuzzy system with BP neural network, use series combination. First, status signal is fuzzy, the fuzzy signal input to the neural network, composed of fuzzy neural network fault diagnosis method.BP neural network connection weights are optimized by using genetic algorithm, and the network is retrained with the optimized weights and thresholds which are used as the initial neural network weights and thresholds of the network. Simulation results show that, GA-BP network is better than traditional BP neural network;Completed the elevator remote monitoring and fault diagnosis system online design and implementation. Results of this study to other similar systems have a certain reference value. |
PDF Full Text Request