模拟电路中黏菌算法优化ELM故障诊断模型研究

Research on slime mould algorithm optimized ELM fault diagnosis model in analog circuit

模拟电路中故障信息复杂多样,为提高故障诊断准确率,提出一种黏菌算法(SMA)优化极限学习机(ELM)的模型。首先,采用线性判别分析(LDA)方法对故障电路原始数据集进行降维,得到ELM网络训练所需的数据;其次,针对ELM随机生成的输入权值和隐含层偏置易导致模型泛化能力差的问题,使用SMA优化ELM的输入权值和隐含层偏置,以获得更优、更稳定的ELM网络参数,提高故障诊断能力。连续可变状态(CTSV)滤波器电路和Sallen-Key带通滤波器的诊断实例表明,SMA优化ELM的故障诊断模型提升了ELM模型的分类效果,具有更优的故障诊断性能。

The fault information in analog cirouit is complex and various,in order to improve the accuracy of fault diagnosis,a model of slime mould algorithm(SMA)to optimize extreme learning machine(ELM)is proposed.Firstly,linear discriminant analysis(LDA)is used to reduce the dimension of the original data set of fault circuit to obtain the data needed for ELM network training.Secondly,aiming at the problem that the input weights and hidden layer bias generated randomly by ELM easily lead to poor generalization ability of the model,SMA is used to optimize the input weights and hidden layer bias of ELM to obtain better and more stable network parameters of ELM and improve fault diagnosis ability.The diagnosis examples of continuous-time state-variable(CTSV)filter circuit and Sallen-Key band-pass filter show that SMA optimized ELM fault diagnosis model improves the classification effect of ELM model and has better fault diagnosis performance.