A physics-based degradation modeling framework for diagnostic and prognostic studies in electrolytic capacitors

Abstract

Avionics systems play a critical role in many aspects of aircraft flight control. As the complexity of these systems increase, the chances of in-flight malfunctions are also likely to increase. This drives the need for Integrated Vehicle Health Management (IVHM) technologies for flight-critical avionics. Studying and analyzing the performance degradation of embedded electronics in the aircraft domain will help to increase aircraft reliability, assure in-flight performance, and reduce maintenance costs. Further, an understanding of how components degrade as well as the capability to anticipate failures and predict the remaining useful life (RUL) can provide a framework for condition-based maintenance. To support a condition-based maintenance and a safety-critical analysis framework, this thesis conducts a detailed study of the degradation mechanisms of electrolytic capacitors, an important component of most electronic systems.

Electrolytic capacitors are known to have lower reliability than other electronic components that are used in power supplies of avionics equipment and electrical drivers of electro-mechanical actuators of control surfaces. Therefore, condition-based health assessment that leverages the knowledge of the device physics to model the degradation process can provide a generalized approach to predict remaining useful life as a function of current state of health and anticipated future operational and environmental conditions.

We adopt a combined model and data-driven (experimental studies) approach to develop physics-based degradation modeling schemes for electrolytic capacitors. This approach provides a framework for tracking degradation and developing dynamic models to estimate the RUL of capacitors. The prognostics and RUL methodologies are based on a Bayesian tracking framework using the Kalman filter and Unscented Kalman filter approaches.

The thesis makes contributions to physics-based modeling and a model-based prognostics methodology for electrolytic capacitors. Results discuss prognostics performance metrics like the median relative accuracy and the á-ë (alpha-lambda) accuracy. We have also demonstrated the derived physics-based degradation model is general, and applied to both accelerated and nominal degradation phenomena. Our overall results are accurate and robust, and, therefore, they can form the basis for condition-based maintenance and performance-based evaluation of complex systems.