An intelligent thermal comfort control strategy for air conditioning of fuel cell vehicles

Abstract

Thermal comfort is an important indicator for evaluating vehicle air conditioning, with many influencing factors, but traditional vehicle air conditioning control only take temperature as the control target, neglecting the passenger's actual thermal comfort situation, which cannot automatically adapt to the thermal comfort needs of different passengers. To address this issue, this paper establishes an air conditioning model based on a fuel cell vehicle (FCV) and proposes a control strategy based on thermal comfort, using deep reinforcement learning based on PPO algorithm to establish the control model. Which can automatically control the compressor and blower, this paper conducts simulation analysis under four different environmental conditions which including static and dynamic situations. The results show that when using the intelligent thermal comfort control strategy, the overall thermal sensation values of passengers are 0.006, 0.005, 0.033, and 0.102, respectively; which are 0.088, 0.333, 0.093, and 0.120 lower than the temperature-based control strategy, this proves that the control strategy proposed in this paper can achieve better thermal comfort for passengers.