IECMAC parameter identification-based power electronic circuit failure predicting method

Abstract

The invention discloses an IECMAC parameter identification-based power electronic circuit failure predicting method, belonging to the technical field of power electronic failure test. The method comprises the following steps of: taking a set power device parameter value as an input training minimum neural network to obtain a failure sample set; judging the failure modal of an electronic circuit to be tested using information entropy-added detecting point electric signal training minimum neural network, and comparing the future time circuit performance parameter predicted by the SVR (support vector regression) with a circuit health threshold, to judge the failure modal of the electronic circuit to be tested; and predicting the real-time failure modal of the power electronic circuit to be tested by taking each detecting node electric signal which is timely monitored as an input repetition training minimum neural network. According to the method, a mathematical model of complex nonlinear system can not be built, and the generalization capability of an identifier can be further improved; and the regression can be carried out on the time sequence formed by system performance parameters by the SVR (support vector regression), the predicting precision and efficiency can be improved, and the on-line and real-time failure prediction can be realized.

Description

technical field [0001] The invention discloses a power electronic circuit fault prediction method based on IECMAC parameter identification, and belongs to the technical field of power electronic fault testing. Background technique [0002] The power circuit in the airborne power electronic device is composed of power components. When placed in a certain space environment, its components are not only subjected to high-frequency start / stop operations, overvoltage, and overcurrent operations, but their performance is also very easy. Affected by external mechanical pressure (shock), EMI, ambient temperature / humidity, salinity and other stresses, which may cause changes in device parameters, if the changes are greater than the allowable range, it will often lead to circuit output performance Deterioration (for example: waveform quality deterioration, THD increase, etc.), and even cause output function failure, which seriously threatens flight safety and ultimately affects the smo...

AI Technical Summary

Problems solved by technology

The advantage of this type of method is that it does not require an accurate mathematical model of the object, and at the same time can effectively express the empirical knowledge of the relevant experts of the object, but it is powerless for specific quantitative analysis

Images