A Control Strategy With Hybrid Compressor Modeling for Electric Vehicle CO 2 Heat Pump Systems

Abstract

To address the significant range attenuation of electric vehicles in cold climates, this study proposes a simulation and optimization framework for a transcritical CO 2 heat pump system. Given the computational inaccuracies of traditional polynomial models in characterizing compressor performance under supercritical conditions, a hybrid compressor sub‐model is developed. This sub‐model integrates a physical foundation with a Particle Swarm Optimization‐Back Propagation (PSO‐BP) neural network to accurately predict mass flow rate and power consumption, achieving a mass‐flow‐rate MAPE within 1.5% on the testing set, with power consumption errors within 0.5%–3.89% and mass flow rate errors within 0.5%–4.71% over the full operating envelope. This high‐accuracy compressor model is then embedded within a system‐level physical model built in MATLAB/Simscape. Based on this enhanced simulation platform, a Phased Variable PTC Cooperative Control Strategy is proposed to resolve the conflict between rapid heating and energy efficiency. Simulation results demonstrate that at −25°C, the proposed strategy reduces cabin warm‐up time by 57.1% compared with the pure heat pump mode while achieving an average COP of 1.776 over the entire warm‐up process, representing a 77.6% improvement in COP compared with conventional pure PTC heating (COP = 1.000). The research demonstrates that enhancing critical component modeling with data‐driven techniques can effectively support the development of precise, energy‐efficient thermal management strategies for EVs.