粒子群算法优化特征和神经网络的模拟电路故障诊断

Analog circuit fault diagnosis based on particle swarm optimization feature and neural network

模拟电路受到自身特性和外界环境的影响,故障变化具有非线性、时变性,针对当前模拟电路故障诊断模型的特征和分类器参数不匹配的难题,提出一种粒子群算法选择特征和神经网络的模拟电路故障诊断模型。首先对当前模拟电路故障诊断现状进行分析,指出它们存在的缺陷;然后提取模拟电路故障诊断特征,利用神经网络作为模拟电路故障诊断分类器;最后采用粒子群算法对模拟电路故障特征与神经网络参数进行优化,在Matlab 2012平台进行了仿真实验。结果表明,该模型的模拟电路故障诊断性能要远远优于其他参比模型,具有广泛的应用前景。

The analog circuit is influenced by its characteristics and external environment, and its fault is non-linear and time-varying. The available fault diagnosis models of analog circuit are difficult to solve the match problem of features and classi- fier parameters, an analog circuit fault diagnosis model based on particle swarm algorithm optimizing feature and neural network is presented. The current situations of analog circuit fault diagnosis are analyzed, and their shortcomings are pointed out. The features of analog circuit fault diagnosis are extracted. The neural network is used as the classifier of analog circuit fault diagno- sis. The analog circuit fault features and neural network parameters are optimized with particle swarm optimization, and simulated with Matlab 2012. The results show that the performance of the proposed model is superior to that of other reference models, and has wide application prospects.