Model predictive control for ice-storage air conditioning systems with time delay compensation integration

Under the optimized control of building demand response, the ice storage air conditioning system can mitigate peak electrical grid pressure and intermittent renewable energy integration issues. However, due to the complex phase-change characteristics of the ice melting and cooling process, such as nonlinearity and significant inertia, current heat transfer and control models fail to optimize the dynamic response performance of the equipment considering these time-delay properties. In this study, an enthalpy-based model is employed to piecewise linearize the nonlinear phase-change process of the ice storage tank. By utilizing the state-space approach to transform the differential equations into transfer function-based dynamic heat transfer equations, the dynamic delay times for the ice storage system are characterized from inertia and flow delay times. Thereby, a time-delay compensation module is embedded into the Model Predictive Control (MPC). Then a dual-objective operational control strategy considering both cost-effectiveness and dynamic response performance is proposed. The results show that, compared to traditional Proportional-Integral-Derivative (PID) controllers, the MPC strategy respectively reduces the response time during the initial, middle, and final phases of ice melting by 43.3 %, 47.1 %, and 50.5 % and contributes to a reduction of peak electrical load and operating costs by 6.5 % and 8.5 %.