基于简化GFRF模型辨识算法的非线性模拟电路故障诊断

Fault Diagnosis in Nonlinear Analog Circuit Based on GFRF Simplified Model Identification Algorithm

提出了基于简化GFRF(广义频率响应函数)模型辨识非线性电路的故障诊断方法。利用非线性频谱分析法获得被诊断系统的GFRF模型估计并提取能表征系统工作状况的电路特征,通过预设故障模式确定被诊断系统所处的工作状态从而定位系统故障。实验结果表明:基于非线性GFRF分析的故障诊断技术,无需在线辨识GFRF、计算量小,故障诊断效果较好。

Fault diagnosis method based on GFRF （Generalized Frequency Response Function ） simplified model to identify nonlinear analog circuit was presented. The GFRF model estimation about diagnostic system can be obtained by nonlinear frequency spectrum analysis, then, the circuit features as the representation of work states in system were extracted from GFRF model, the faults in diagnostic system can be located by pre-set fault mode. In this fault diagnosis method, the nonlinear system identification and feature exteaction were realized by simplified GFRF model. The fault diagnosis technology based on nonlinear GFRF analysis avoided GFRF identification online and realized easily. An illustration validates the better diagnostic effect.