基于IBPSO的模拟电路故障特征提取方法

Fault Feature Extraction Method of Analog Circuits Based on IBPSO

针对有效采样点法提取故障特征时冗余信息多、易造成维数灾等问题,提出利用改进的二进制粒子群算法提取故障特征。研究粒子群优化算法和二进制粒子群优化算法的差异以及在故障特征提取方面存在的不足,通过改进群体极值的更新方式避免搜索结果陷入局部最优。以Sallen-Key带通滤波器为诊断实例,完成9类模拟电路故障模式的特征提取。结果表明:通过该方法进行特征提取可有效降低故障诊断模型的复杂性,与二进制粒子群优化算法相比,该方法在特征维度和诊断准确率上具有明显的优势。

According to the problems such as redundancy information and dimension disaster caused by fault feature extraction of effective sampling point method, put forwards improved binary particle swarm method to extract fault feature. The differences and deficiencies between particle swarm optimization （PSO） algorithm and binary particle swarm optimization （BPSO） algorithm are analyzed. Change swarm extremum update mode to avoid research result to get into local best. Taking Sallen-Key filter as example, realize feature extraction of 9 simulation circuit fault mode. The result shows that the proposed method can effectively reduce the complexity of the fault diagnosis model. Compared with the binary particle swarm optimization algorithm, it has obvious advantages in the feature dimension and diagnostic accuracy.