摘要
提出了一种小波包分析与最小二乘支持向量机相结合的机械设备故障诊断模型。首先对故障信号功率谱进行小波分解,简化了故障特征向量的提取,然后采用最小二乘支持向量机进行故障诊断。在该模型中,用二次损失函数取代支持向量机中的不敏感损失函数,将不等式约束条件变为等式约束,从而将二次规划问题转换为线性方程组的求解,并提出对核函数的σ参数进行动态选取。仿真结果表明:该模型可以取得较高的故障诊断准确率。
A machinery fault diagnosis model combining the wavelet packet analysis and least squares support vector machine （LSSVM） together was presented. First the power spectrum of fault signals were decomposed by wavelet analysis, which predigests choosing method of fault eigenvectors, and then the LSSVM was adopted to diagnose faults. In the model, the non-sensitive loss function was replaced by quadratic loss function and the inequality constraints were replaced by equality constraints. Consequently, quadratic programming problem was simplified as the problem of solving linear equation groups. It was presented to choose σ parameter of kernel function on dynamic. The simulation results show the model enhances preciseness rate of diagnosis.
出处
《计算机应用研究》
CSCD
北大核心
2007年第7期99-101,共3页
Application Research of Computers
基金
中国博士后科学基金资助项目(2005038515)
关键词
小波包分析
故障诊断
特征向量
最小二乘支持向量机
核函数
wavelet packet analysis
fault diagnosis
eigenvector
least squares support vector machine
kernel function
