基于小波包和RBF神经网络的瓦斯传感器故障诊断

针对瓦斯传感器故障诊断速度慢、诊断精度不高的问题，以常见的冲击型、漂移型、偏置型和周期型传感器输出故障为研究对象，提出了一种基于减聚类( SCM)与粒子群( PSO)算法优化的RBF神经网络进行模式分类与辨识的瓦斯传感器故障诊断方法。首先，利用三层小波包分解得到各个节点的分解系数，采用一定的削减算法使故障的瞬态信号特征得到加强，获取最优的特征能量谱。再利用SCM￣PSO算法优化RBF神经网络，使粒子的搜索速度更快，更有利于发现全局最优解。最后通过实验对比分析，该方法具有训练速度快、分类精度高的特点，辨识正确率在95％以上，能够显著提高故障诊断的速度和准确性。

Gas sensor fault diagnosis based on wavelet packet and RBF neural network identification

In order to solve the problem that the gas sensor diagnosis speed is slow and the diagnosis accuracy is not high,this paper takes the common type gas sensor fault such as impact,drift,bias and periodic fault as research ob￣ject,and proposes a pattern classification and identification of the fault diagnosis of gas sensor method based on RBF neural network that is optimized by subtractive clustering and particle swarm optimization algorithm. Use three layer wavelet packet decomposition technologies to get the coefficients of each node,and adopt some cutting algorithm to improve the transient signal features of the fault,and then obtain the optimal energy spectrum. Next,use SCM￣PSO algorithm to optimize RBF neural network and make the particles search faster and easier to find the global optimal solution. Finally,through experiment contrast analysis,this method has the features of fast training speed,high accu￣racy of classification,and the identification correct rate is more than 95%. It can significantly improve the effective￣ness and accuracy of the fault diagnosis.