Real-time Torque Distribution Simulation of Parallel Hybrid Vehicle Engine

Jing Wang^*

• Puyang Vocational and Technical College, Puyang, China

Parallel hybrid vehicles are a new type of automobile developed to address environmental pollution and the energy crisis. However, the current engine torque distribution technology suffers from slow feedback speed and poor distribution effectiveness. Therefore, this study proposes an optimal strategy for torque distribution based on the Markov Decision Process framework using a Proximal Policy Optimization algorithm. The model also integrates Fiber Bragg Grating sensors to construct a real-time torque distribution system. In the algorithm performance test, the accuracy and loss rate at the end of the iteration were 93.2% and 1.0%, respectively, while the average feedback time of the algorithm was only 32 ms. In the model simulation experiment, the time for real-time torque distribution did not exceed 70 ms, and the energy utilization rates during startup, normal driving, acceleration, deceleration braking, and high-speed driving were 75.5%, 42.3%, 41.5%, 22.5%, and 50.0%, respectively. In the robustness test, the model maintained an accuracy of 82.3% under environmental noise interference of 300 dB and 83.1% at an ambient temperature of 180°C. The above results demonstrate that the proposed model offers high accuracy and robustness, enabling efficient utilization and recovery of energy in parallel hybrid vehicle engines through real-time torque distribution. This provides a new approach to improving energy utilization efficiency and contributes to solving environmental pollution and the energy crisis.