Multi-evaporator air conditioning system has found increasing applications in buildings. Energy efficient control for this kind of system is a critical issue. The major challenge is the complicated coupling in operation parameters and strong nonlinearities. This paper proposes a novel energy-efficient decentralized control strategy for multi-evaporator air conditioning systems. In this strategy, the compressor speed is controlled to maintain the suction superheat, and the indoor expansion valve is controlled to maintain room temperature of each indoor unit. The command following test shows that the proposed strategy delivers more stable and smooth control performance compared with a conventional strategy. Moreover, simulation results show that the proposed strategy provides 8.6% energy savings. The proposed strategy is also found to be more robust against model error and actuator faults, compared with a model predictive controller.
Keywords multi-evaporator, air conditioning system, decentralized control, energy saving, refrigeration