面向SiC MOSFET阵列的均衡控制方法、系统、电子设备及介质
By combining thin-film thermocouple arrays and convolutional neural networks, temperature data of SiC MOSFET arrays are acquired and analyzed in real time, and gate drive signals are dynamically adjusted. This solves the problems of insufficient delay and accuracy in the equalization control of SiC MOSFET arrays, and achieves efficient device equalization and thermal runaway risk assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SPEYI TECH (BEIJING) CO LTD
- Filing Date
- 2026-04-17
- Publication Date
- 2026-07-17
AI Technical Summary
Existing equalization control schemes for SiC MOSFET arrays suffer from drawbacks such as centralized controllers processing massive amounts of data, resulting in delays, difficulty in adapting to dynamic changes in devices, and the possibility that adjustment strategies based on a single threshold model may amplify current deviations. Furthermore, the lack of hierarchical assessment of thermal runaway risk leads to insufficient control accuracy.
Temperature data is acquired in real time by a thin-film thermocouple array, and a three-dimensional temperature field is constructed by combining it with a thermal characteristic curve database. Key information is extracted using a convolutional neural network, and a risk index is output. Distributed balanced control is achieved by dynamically adjusting the duty cycle of the gate drive signal through an intelligent agent.
It achieves precise and timely balanced control of SiC MOSFET arrays, avoiding local imbalance and thermal runaway, and ensuring high reliability and long lifespan of the equipment.
Smart Images
Figure CN122052508B_ABST