Application of a multivariable adaptive control strategy to automotive air conditioning systems

This paper presents the application of a multivariable adaptive control strategy to a typical automotive air conditioning system. An experimentally validated physical model for the air conditioning (a/c) cycle is first presented and is subsequently used to choose a relevant model structure for indirect adaptive control. Recursive identification of this model structure is carried out using a multi-input multi-output (MIMO) parameter estimation algorithm to obtain an equivalent discrete-time state space model of the a/c system. Linear quadratic regulator (LQR) design is implemented on the estimated model with the objectives of reference tracking and disturbance rejection. Simulation studies are presented to evaluate the advantages of using the electronic expansion valve and the air flow rate over the evaporator to control the efficiency and the capacity of a general automotive a/c unit using this adaptive control approach. The results demonstrate the efficacy of the MIMO controller and motivate further research in this area.