Long-Term Prediction of Multistress Accelerated Aging of Capacitors by Long Short-Term Memory Network
- Resource Type
- Periodical
- Authors
- Liu, H.; Claeys, T.; Pissoort, D.; Vandenbosch, G.A.E.
- Source
- IEEE Transactions on Instrumentation and Measurement IEEE Trans. Instrum. Meas. Instrumentation and Measurement, IEEE Transactions on. 72:1-6 2023
- Subject
- Power, Energy and Industry Applications
Components, Circuits, Devices and Systems
Capacitors
Logic gates
Neurons
Computer architecture
Capacitance
Stress
Prediction algorithms
Aging prediction
capacitance
capacitor aging
equivalent series resistance (ESR)
long short-term memory (LSTM)
- Language
- ISSN
- 0018-9456
1557-9662
To forecast the long-term aging deviations of capacitors, a new highly performing deep neural network (DNN) architecture is proposed based upon the long short-term memory (LSTM) algorithm. By importing the early aging data with respect to the applied thermal and electrical stresses into the proposed LSTM-based DNN architecture, the future accelerated aging-induced deviations in capacitance and equivalent series resistance (ESR) are predicted accordingly. The dropout and the prediction interval (PI) technique are applied to overcome overfitting issues and obtain an uncertainty estimation. The results indicate that the proposed LSTM-based DNN algorithm has a higher prediction accuracy and narrower PIs compared with other deep learning methods.