Correntropy-based Aligned Predictor for Degradation Prognosis
- Resource Type
- Conference
- Authors
- Mei, Wenjuan; Su, Yuanzhang; Liu, Zhen
- Source
- 2021 CAA Symposium on Fault Detection, Supervision, and Safety for Technical Processes (SAFEPROCESS) Fault Detection, Supervision, and Safety for Technical Processes (SAFEPROCESS), 2021 CAA Symposium on. :1-6 Dec, 2021
- Subject
- Components, Circuits, Devices and Systems
Robotics and Control Systems
Signal Processing and Analysis
Degradation
Insulated gate bipolar transistors
Training
Estimation
Market research
Robustness
Safety
Condition Monitoring
Extreme Learning Machine
Correntropy
- Language
With large usage of mechanical and electric systems, condition monitoring and prognosis have urgently required improvements to meet increasing complexity and variable operating conditions. However, existed methods can hardly obtain high information usage and keep robustness considering the non-identity of objects. Considering both similarity and non-identity of degradation observations, we propose the correntropy-based aligned predictor (CAP) to estimate the degradation condition of different devices with the same type. We use the subspace clustering to provide an aligned matrix to describe the relationships between objectives and develop the ensemble learning technique with correntropy-based extreme predictor. The degradation prediction of IGBTs’ accelerated aging testing proves that CAP enhances the information usage during degradation prediction and achieve high robustness.