Gradual fault diagnosis and life prediction method of uncertain sophisticated circuit system

Abstract

A gradual fault diagnosis and life prediction method of an uncertain sophisticated circuit system is provided. The method comprises the following steps: S1, performing modeling on the uncertain sophisticated circuit system, to obtain an uncertainty global diagnosis sophisticated bond graph model; S2, obtaining augmented global analytical redundancy relations AGARRs, obtaining a global adaptive threshold by using the augmented global analytical redundancy relations AGARRs, and performing fault detection on the sophisticated circuit system when a parameter is uncertain; S3, analyzing structuralcharacteristics of the augmented global analytical redundancy relations AGARRs, to provide a global fault characteristic matrix of the system, to isolate a possible fault on this basis; and S4, approaching the gradual fault by using a degradation model of a time-varying attenuation rate, estimating a fault value and a fault occurrence time by using an enhanced expansion Kalman filtering algorithm,calculating a gradual degradation model coefficient, and predicting a remaining useful life of a fault element. The foregoing method can be used to implement fault diagnosis and remaining useful lifeprediction.

Description

technical field [0001] The invention relates to the field of model-based gradual fault diagnosis and remaining service life prediction of an uncertain hybrid system, in particular to a method for gradual fault diagnosis and life prediction of an uncertain hybrid circuit system. Background technique [0002] Uncertain mixed circuit systems are used in all walks of life. If one or some components in the circuit system fail, the entire system will fail. At this time, equipment maintenance personnel need to make timely judgments on the equipment, or It can effectively prevent the service life of components. However, the degradation of components in the system is gradual, which makes it difficult to detect and eliminate early faults in time. [0003] The closest prior art of this application is the publication "A Fault Diagnosis and Estimation Method for Electric Scooter Based on Bond Graph Model". Although the fault diagnosis and estimation method is disclosed, the real-time fa...

AI Technical Summary

Problems solved by technology

However, the degradation of the components in the system is gradual, which makes it difficult to realize early failure and detect and eliminate it in time

[0003] The closest prior art of this application is the publication "A Fault Diagnosis and Estimation Method for Electric Scooter Based on Bond Graph Model". Although the fault diagnosis and estimation method is disclosed, the real-time fault is not disclosed. Parameter Estimation and Remaining Service Life Prediction Methods

Images