Skip to Main content Skip to Navigation

Contribution to condition monitoring of Silicon Carbide MOSFET based Power Module

Abstract : More electrical aircraft requires power modules of higher performances, especially in terms of reliability with a control of lifetime. The replacement of hydraulic and pneumatic systems by electric actuators and their associated converters is the present trend to reduce maintenance cost and fuel consumption. Adding more electric components is also thought as a good way to increase reliability in systems. Reliability is still analysed from accelerated stress cycles. A large volume of data must be obtained in various conditions to assert a pertinent extrapolation of remaining lifetime during operation. A trend is to embed some condition monitoring functions in power modules to help predict the remaining lifetime. This approach is the field of hardware developments with respect to sensors and decorrelation methods but mainly dedicated to one particular failure. This thesis presents a learning approach of silicon carbide MOSFET based power modules condition monitoring. A large literature study has led to the elaboration of a test plan and an instrumented test bench. This test bench allows an accelerated lifespan of power module and an on-line recording of several electrical parameters. These parameters shows a drift according to the power module ageing. A neural network model based on these parameters drifts has been constructed to estimate the remaining useful lifetime of a power module in normal operation
Document type :
Complete list of metadata

Cited literature [87 references]  Display  Hide  Download
Contributor : ABES STAR :  Contact
Submitted on : Friday, March 8, 2019 - 11:44:08 AM
Last modification on : Monday, September 13, 2021 - 2:44:04 PM
Long-term archiving on: : Monday, June 10, 2019 - 4:21:04 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02061648, version 1


Malorie Hologne. Contribution to condition monitoring of Silicon Carbide MOSFET based Power Module. Electric power. Université de Lyon, 2018. English. ⟨NNT : 2018LYSE1317⟩. ⟨tel-02061648⟩



Record views


Files downloads