A Rule-Based Control for Fuel-Efficient Automotive Air Conditioning Systems

In a conventional passenger vehicle, the AC system is the largest ancillary load. This paper proposes a novel control strategy to reduce the energy consumption of the air conditioning system of a conventional passenger car. The problem of reducing the parasitic load of the AC system is first approached as a multi-objective optimization problem. Starting from a validated control-oriented model of an automotive AC system, an optimization problem is formalized to achieve the best possible fuel economy over a regulatory driving cycle, while guaranteeing the passenger comfort in terms of cabin temperature and reduce the wear of the components. To complete the formulation of the problem, a set of constraints on the pressure in the heat exchanger are defined to guarantee the safe operation of the system. The Dynamic Programming (DP), a numerical optimization technique, is then used to obtain the optimal solution in form of a control sequence over a prescribed driving cycle. The solution of the DP is analyzed with the scope of understanding the system optimal behavior and extracting rules to mimic the response of the DP but in a forward-looking implementation. The analysis of the global optimal solution allows for the synthesis of a sub-optimal rule-base control, which is then implemented and verified on a test vehicle.