Virtual Unmodeled Dynamics Robust Predictive Control With Time-Delay Compensation for Automobile Air Conditioning Process

In the context of new energy vehicles, it is common for the compressor to be directly powered by an electric motor. However, the demanding operational circumstances frequently lead to substantial variations in motor speed. This work presents a novel approach for the speed control system of air conditioning compressors, which involves the utilization of a virtual unmodeled dynamics robust predictive control with time-delay compensation. The initial step involves analyzing the conversion mechanism between the angular velocity of the compressor and the duty ratio, while considering the delay effects resulting from inertia. Subsequently, a state space model is constructed, incorporating unmodeled dynamics. In addition, in light of the uncertain attributes of the unmodeled dynamics, a signal processing function is formulated for the unmodeled dynamics through a time compensation data strategy and an interval differential compensation strategy. Furthermore, a robust predictive controller with time delay and unmodeled dynamics compensation techniques is presented. This controller integrates an unmodeled dynamics time compensation unit and a differential compensation unit, utilizing adaptive backstepping control design techniques. Namely, the control signal is compensated using the unmodeled dynamics signal processing function. The effectiveness and feasibility of the suggested technique are evaluated by simulation and real operation on an air conditioning system platform.