摘要
论文提出一种基于BP和WTA神经网络的轴承故障诊断模式识别新方法,将轴承故障数据提取为所需的特征向量,训练BP-WTA,得到基于BP和WTA神经网络的诊断分类器。通过144组样本进行分类识别轴承损伤程度实验,基于BP和WTA神经网络方法与传统BP、HMM方法进行对比,其诊断准确率达到100%,验证了该方法在轴承故障诊断中的有效性和实用性。
This paper presents a novel pattern recognition method of fault diagnosis of rolling bearing which is based on BP and WTA neural networks.In the course of fault diagnosis,data on rolling bearing fault is transformed into desired feature vector to input to train BP-WTA,and diagnostic classification of BP and WTA neural network is obtained.By 144 experimental group samples to classify the degree of bearing damage,the diagnostic accuracy is expected to be 100%compared the BP and WTA Neural Networks with traditional method of BP and HMM,which shows that the proposed method is effective and practical in bearing fault diagnosis.
出处
《计算机与数字工程》
2017年第2期291-298,共8页
Computer & Digital Engineering
基金
国家自然科学基金(编号:61370202
61550110248)
中央高校基本业务费(编号:ZYGX2013J041)资助