摘要
针对传统方法难以精确检测风力发电机组齿轮箱非线性、非平稳振动信号以及现有许多故障诊断方法无法有效诊断齿轮箱早期故障的问题,首先引入排列熵算法对齿轮箱振动信号进行早期故障分析,进而引入多尺度排列熵算法实现原始振动信号的特征提取,得到故障诊断的样本数据,最后将其输入到建立的基于遗传算法优化支持向量机的诊断模型中,完成故障模式的识别与分类。仿真结果表明,该方法能够有效识别齿轮箱的异常工况,具有较高的故障诊断精度。
Aiming at the problem that with the traditional method it is difficult to accurately detect the non-linear and non-stationary vibration signals of the wind turbine＇s gearbox and the existing fault diagnosis methods can not effectively di- agnose the early fault of the gearbox, this paper firstly intro- duces the permutation entropy algorithm to analyze the early fault of the gearbox vibration signal and then the multi-dimen- sion permutation entropy algorithm to realize the feature extraction of the original vibration signal, so as to get the sample data of the fault diagnosis. Finally, the sample data is input into the diagnosis model based on genetic algorithm optimization support vector machine to complete the fault pattern recognition and diagnosis. The simulation results show that the method can effectively identify the abnormal working conditions of the gear box with high fault diagnosis accuracy.
出处
《电网与清洁能源》
北大核心
2017年第5期87-91,103,共6页
Power System and Clean Energy
基金
国家电网科技项目(522722150012)
陕西水利科技计划项目(2015s1kj-04)~~
关键词
齿轮箱
多尺度排列熵
遗传算法
支持向量机
故障诊断
gearbox
multi-dimension permutation entropy
genetic algorithm
support vector machine
fault diagnosis
作者简介
刘骊（1990-），男，硕士研究生，研究方向为电力系统新能源；
贾嵘（1971-），男，博士，教授，博士生导师，主要研究方向为电力系统自动化。
