Abstract:

The Electric Vehicle Thermal Management (EVTM) system's temperature variation is a key factor affecting energy consumption. To optimize energy strategies, a good temperature prediction method is crucial. Factors like ambient temperature, vehicle speed, and air conditioning compressor speed influence temperature variation. This paper proposes a method using Particle Swarm Optimization (PSO) and Back propagation Neural Network (BP) to predict temperature. Real-time data from road experiments is used to ensure accuracy. Simulation results show that the PSO-BP method reduces prediction errors by 66%, 75%, and 25% compared to the BP neural network.