This work presents the benefits of using a model predictive control (MPC) approach for controlling a high efficiency absorption chiller-based solar cooling system with thermal energy storage, incorporating perfect solar resource and load forecasting information. A dynamic physics-based model of the solar air-conditioning system has been built for studying the system behavior. A genetic algorithm based predictive controller is utilized to minimize backup energy consumption while satisfying the cooling demand. The simulations have been carried out using the open-source programming language Python. Detailed investigation of the role of the predictive controller and its decision strategy have been carried out using ten and fifty days simulations. Effect of storage tank heat losses has been investigated. For the simulated example case pertaining to a building, results show the model predictive controller usage delivers about 10% reduction in auxiliary energy use in the system. This is achieved through reduction in tank heat losses, better utilization of heat stored in the tank. It is seen that the MPC based controller enables new system operational capabilities by running the solar collector pump in variable flow mode and allowing the simultaneous heat delivery from storage and backup devices. Opportunities to improve the MPC benefits have been identified. The benefits of the MPC are seen to be sensitive to the system parameters and specific constraints. In summary, this paper provides valuable insights into solar cooling system design and control.
