AI-Driven Two-Phase R134a Thermal Management and Intelligent Control of Si MOSFET Traction Inverters for Next-Generation Electric Vehicles

Compared with conventional silicon-based Insulated Gate Bipolar Transistors (IGBTs), Silicon Metal–Oxide–Semiconductor Field-Effect Transistor (Si-MOSFETs) offer superior performance characteristics, including higher operating temperature capability, faster switching speeds, and higher switching frequencies. As a result, Si MOSFETs are widely regarded as a key enabling technology for next-generation electric drive systems. Their adoption in electrified vehicles has led to significant improvements in power conversion efficiency, power density, and thermal management, while also reducing the size and complexity of cooling systems. Current inverter models globally face critical challenges in delivering real-time responsiveness for autonomous drivetrains, preventing temperature spikes, reducing switching losses, and avoiding vehicle shutdowns even when supported by a 24 V DC auxiliary power supply. The aim of the project is to develop an AI-based R134a two-phase thermal management system (ThMS) Si MOSFET inverter for the EVs/AEVs to overcome the constraints of the conventional EV’s inverter system. The proposed system will be developed by integrating scalable AI-assisted control algorithms, IoT-based monitoring and control architecture, highly efficient MOSFET PWM control modulation, a 2-phase R134aThMS, and an intuitive auto-diagnostic user interface. By introducing a closed-loop AI Si MOSFET PWM method and active two-phase cooling with a R134a refrigerant, the proposed system allows the inverter to dynamically respond to EV power demands while controlling thermal loads in the range of 25–50 °C in milliseconds.