基于模糊神经网络的电梯实时监控和在线故障诊断的研究

发布时间：2018-01-21 01:39

本文关键词： 电梯故障诊断 BP神经网络模糊系统遗传算法　出处：《东北大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着现代化城市的快速发展,电梯已成为不可缺少的配套工具。电梯的安全性直接关系到人身安全,及时发现故障和及时检修故障是确保电梯可靠运行、提高电梯监控技术水平和保证人身安全的关键。电梯控制系统在交流变频变压调速电梯系统中是最重要的部分,也是电梯系统中故障频发的系统,目前电梯故障诊断方法一般是依靠技术人员的感觉和经验,难以确保故障诊断的快速性和精准性。针对这一现状,本文展开了电梯控制系统故障诊断的研究,目的在于当电梯发生故障时能及时诊断出故障,同时能及时采取相应的措施排除故障,保证人员和设备的安全。首先,整个电梯系统及其复杂,所采集的状态参数非常庞大,本文只对电梯控制系统进行研究。电梯控制系统故障信号的不确定性、模糊性及非线性,同时考虑到采集数据具有周期性,各个参数的量纲也不同,本文将模糊系统和BP神经网络进行结合的办法,采用串联结合方式；先将状态信号进行模糊化处理,将模糊化后的信号输入到神经网络,组成模糊神经网络的故障诊断方法。利用遗传算法优化BP神经网络的连接权值和阈值,用优化得到的权值和阈值作为神经网络的初始权值和阈值对网络重新进行训练,仿真结果表明遗传算法优化的BP网络明显优于传统的BP网络。最后完成了电梯远程监控与故障在线诊断系统的设计与实现。本文研究成果对其他类似系统有一定的参考价值。
[Abstract]:With the rapid development of modern cities, elevators have become an indispensable supporting tool. The safety of elevators is directly related to personal safety. Timely detection of faults and timely maintenance of faults is to ensure the reliable operation of elevators. Improve the elevator monitoring technology and ensure the key to personal safety. Elevator control system is the most important part of AC variable-frequency variable-voltage adjustable speed elevator system, and it is also the frequently occurring system in elevator system. At present, the method of elevator fault diagnosis generally depends on the feeling and experience of technicians. It is difficult to ensure the rapidity and accuracy of the fault diagnosis. In view of this situation, the research of elevator control system fault diagnosis is carried out in this paper, the purpose is to diagnose the fault in time when the elevator fault occurs. At the same time, can take timely measures to troubleshoot, to ensure the safety of personnel and equipment. First of all, the whole elevator system and its complexity, the collected state parameters are very large, this paper only study the elevator control system. Elevator control system fault signal uncertainty, fuzziness and nonlinearity. At the same time, considering the periodicity of the collected data and the different dimensions of each parameter, the method of combining fuzzy system with BP neural network is adopted in this paper. Firstly, the state signal is processed by fuzzification, and the fuzzy signal is input into the neural network to form the fault diagnosis method of the fuzzy neural network. Genetic algorithm is used to optimize the connection weight and threshold of BP neural network, and the optimized weights and thresholds are used as the initial weights and thresholds of the neural network to re-train the network. The simulation results show that the BP network optimized by genetic algorithm is obviously superior to the traditional BP network. Finally, the design and implementation of elevator remote monitoring and fault online diagnosis system are completed. The research results of this paper have certain reference value to other similar systems.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TU857;TP277

【参考文献】