A novel prediction method about single components of analog circuits based on complex field modeling

Area of Science:

• Electrical Engineering
• Reliability Engineering
• Predictive Maintenance

Background:

• Existing analog circuit prediction methods often lack correlation with circuit analysis, leading to irrational fault indicator (FI) calculations and reduced prognostic performance.
• Faults in single components are the most common failure mode in analog circuits, necessitating focused prediction strategies.

Purpose of the Study:

• To propose a novel prediction method for single components in analog circuits using complex field modeling.
• To enhance the rationality of fault indicator (FI) feature set calculation.
• To improve the accuracy of predicting the remaining useful performance (RUP) of analog circuit components.

Main Methods:

• Circuit structure analysis and transfer function calculation to implement complex field modeling.
• Parameter scanning model to analyze the relationship between parameter variation and component degeneration.
• Development of a novel degeneration trend model for analog circuit components based on the FI feature set.
• Particle filter (PF) application for model parameter updating and RUP prediction.

Main Results:

• A more reasonable FI feature set was obtained through complex field modeling and parameter scanning.
• The proposed model demonstrated improved accuracy in predicting the degeneration trend and RUP of analog circuit components.
• Experimental verification confirmed the effectiveness of the developed prediction method.

Conclusions:

• Complex field modeling provides a rational basis for FI feature extraction in analog circuit prognostics.
• The novel prediction method significantly enhances the accuracy of RUP prediction for analog circuit components.
• This research offers a valuable approach for improving the reliability and maintenance strategies of analog circuits.