摘要
对风机齿轮箱轴承故障诊断进行了研究,提出一种基于分形维数和遗传算法支持向量机(GA-SVM)相结合的故障诊断算法。基于常用的时域特征参数作为支持向量机的识别参数,引入分形维数特征参数来提升支持向量机的识别精度。提出了基于遗传算法(GA)的支持向量机参数优化的模型,通过GA的寻优自动获得最优的支持向量机参数。采用某风场的风电机组齿轮箱轴承数据进行故障诊断,实验表明,所提出的GA-SVM模型很好地解决了参数选择的问题,同时基于分形维数的特征参数也提高了风电机组轴承故障的识别准确率。
For wind turbine gearbox bearing fault diagnosis is studied, and a fault diagnosis method based on the fractal dimension and genetic algorithm support vector machine (GA-SVM) is put forward. Based on the commonly used time domain feature parameters as the support vector machine identification parameters, the fractal dimension feature parameters are introduced to enhance the recognition accuracy of support vector machines. The model of support vector machine parameters optimization based on genetic algorithm is proposed, and the optimal support vector machine parameters are obtained by the optimization of GA. Using the gear box bearing data from a wind farm in Zhangjiakou, Hebei province for fault diagnosis. Experimental results show that the proposed model GA-SVM provided a good solution to the parameter selection problem, as well as the characteristic parameters based on fractal dimension also improve the recognition accuracy of wind turbine bearing failure.
出处
《计量学报》
CSCD
北大核心
2018年第1期61-65,共5页
Acta Metrologica Sinica
基金
国家自然科学基金(51475407)
关键词
计量学
轴承故障诊断
风电齿轮箱
分形维数
遗传算法支持向量机
识别准确率
metrology
fault diagnosis of bearing
wind turbine gearbox
fractal dimension
genetic algorithm support vector machine
recognition accuracy