基于INGO-Transformer的模拟电路元件故障预测  

Component Fault Prediction for Analog Circuits Based on INGO-Transformer

作　　者：杜先君[1] 曹磊 DU Xianjun;CAO Lei(College of Electrical and Information Engineering,Lanzhou University of Technology,Lanzhou 730050,China)

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050

出　　处：《火力与指挥控制》2024年第10期158-166,共9页Fire Control & Command Control

摘　　要：针对模拟电路元件易受外部环境影响发生故障、故障特征提取困难、无法准确预测及诊断元件故障等问题,基于Transformer模型提出改进INGO-Transformer方法。采用小波包分解(WPD)对原始数据进行特征提取,使用特征向量之间的三角距离来表征模拟电路中元件的退化状态,使用INGO优化Transformer的训练超参数构建预测模型。以Sallen-Key带通滤波电路与镜像电流源电路为预测实验对象进行故障预测实验,采用MAE与MSE作为故障预测模型评价指标,两组实验电路10次实验平均MAE、MSE结果分别为4.2162e-04、4.1906e-07和0.0017、1.9625e-05。仿真结果表明,所提方法在模拟电路单一元件故障预测中具有较高的准确性与较强的泛化能力。To address the problems that analog circuit components are susceptible to faults caused by external environment,difficult to extract fault features,and unable to accurately predict and diagnose component faults,etc.An improved INGO-Transformer method is proposed based on the Transformer model.First,wavelet packet decomposition(WPD)is used to extract features from the original data.Then,the delta distance between feature vectors is used to characterize the degradation state of components in the simulated circuit.Finally,the prediction model is constructed using the training hyperparameters of INGO Optimized Transformer.The Sallen-Key bandpass filter circuit and the mirror current source circuit are used as the prediction experiments for the fault prediction experiments.The MAE and MSE are used as the evaluation indexes of the fault prediction models,and the average MAE and MSE results of the two groups of experimental circuits are 4.2162e-04,4.1906e-07 and 0.0017,1.9625e-05 for 10 experiments,respectively.The simulation results show that the proposed method has higher accuracy and stronger generalization ability in the prediction of single component faults in analog circuits.

关 键 词：模拟电路 故障预测 小波包分解 TRANSFORMER 优化算法 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置]