摘要
针对风力发电机故障诊断中存在的故障样本不完备的问题,提出将支持向量数据描述(Support Vector Data Description,SVDD)与支持向量机(Support Vector Machine,SVM)相结合的方法处理风力发电机故障诊断问题.根据风力发电机正常状态样本数据和已知故障样本数据建立数据描述模型,并以已知故障样本数据建立SVM分类器模型.采用数据描述模型对测试样本数据进行拒绝和接受处理:被拒绝的样本为未知故障类型;利用SVM分类器模型对被接受的样本进行具体类别诊断.试验结果表明:该方法可以有效处理风力发电机故障诊断中故障样本不完备问题,能较好识别已知故障并对未知故障给出判断.
As to the issue of incomplete fault samples in the fault diagnosis of wind power generator, a method is proposed to deal with the wind power generator fault diagnosis, which combines the Support Vector Data Description (SVDD) with the Support Vector Machine (SVM). According to the data samples of wind power generator in the normal and known fault states, a data description model is built, and a SVM classifier model is built on the basis of the known fault data samples. The test data samples are rejected or accepted by the data description model. The rejected samples are the unknown fault types; the specific categories of accepted samples are diagnosed by the SVM classifier model. The tests show that, the issue of incomplete fault samples can be solved effectively by the method; the method can effectively handle the known faults and give judgment for the unknown faults.
出处
《计算机辅助工程》
2015年第3期67-71,77,共6页
Computer Aided Engineering
基金
国家自然科学基金(61175059)
河北省自然科学基金(F2014205115)
关键词
风力发电机
故障诊断
支持向量数据描述
支持向量机
wind power generator
fault diagnosis
support vector data description
support vector machine