摘要在最近几年里，大型复杂动力设备故障预测的传统方法已经不能跟上发展的需求，因此新的智能方法在传统方法的基础上得到了快速发展，新型的现代故障预测技术不断涌现：神经网络、模糊逻辑、模糊神经网络、遗传算法等都已经在大型设备故障预测领域得到成功应用。随着现代工业的发展，自动化系统的规模越来越大，也越来越复杂，这样就使得系统产生故障的可能性和复杂性变大了，这个时候只靠一种理论或一种方法，不管它是是智能的还是经典的，都很难实现在这么复杂的条件下对大型设备的故障完全、准确、及时地诊断，但多种方法的综合运用，既可以是经典方法与智能方法的结合，也可以是两种或多种智能方法的结合，这就兼顾了实时性和精确度。因此多种方法的有机融合、综合运用这一趋势将成为必然，也将成为电机故障在线预测技术未来发展的主流方向。本文首先说明了电机轴承的一些结构以及应用，并分析了它们的主要特征，研究了它的机理，以及对轴承常见的故障进行阐述，在了解它们的基础上对其原因进行分析。最后本文将建立新的模型来解决轴承故障的复杂难题,通过相关的了解以及比较以上所提到的各种方法的优缺点，我觉得可以将模糊控制与神经网络结合起来，发挥各自的优点，使故障预测的结果更加精准。并通过MATLAB软件进行仿真测试。60263

毕业论文关键词：电机故障  轴承故障  模糊神经网络  

毕业设计说明书（论文）外文摘要

Title     Large motor bearing failure prediction algorithm  simulation 

Abstract

In recent years, a large complex of traditional power equipment failure prediction methods have been unable to keep up with development needs, so the new intelligent method based on traditional methods has been the rapid development of new technologies are emerging modern failure prediction: Neural Networks , fuzzy logic, neural networks, genetic algorithms and so has been in the field of large-scale equipment failure prediction has been successfully applied. With the development of modern industry, the increasing scale automation systems, but also more complex, thus making the system failure probability and complexity becomes large, and this time rely on a theory or a method whether it is intelligent or classic, are difficult to achieve in such a complex condition of large equipment failure complete, accurate and timely diagnosis, but the integrated use of a variety of methods can be both classical methods and intelligent approach combined, it can be two or more intelligent combination of methods, which take into account the timeliness and accuracy. Therefore, the organic integration of a variety of methods, the integrated use of this trend will become inevitable, will also become online motor failure prediction of future development direction of the mainstream. This paper describes some structural and motor bearing applications, and analysis of their main characteristics, study its mechanism, as well as elaborate bearing common failure in understanding the basis of their analysis of its causes. Finally, we will create a new model to solve complex problems bearing failure, through the relevant knowledge and to compare the above mentioned advantages and disadvantages of the various methods, I think we can be fuzzy control and neural network combined to play their respective advantages, make the results more accurate failure prediction. By MATLAB software simulation testing.

Keywords  Motor failure  Bearing failure  Fuzzy Neural Network

目   次

1  绪论 1

1.1  研究背景和意义 1

1.2  大型电机故障预测国内外研究现状