Solar photovoltaic and ice storage air-conditioning are both green air-conditioning technologies that can relieve the power supply pressure on power grids and reduce the operation cost of air-conditioning system. This study combines these two technologies and proposes a photovoltaic ice storage air-conditioning (PSAC) system that has a charging-and-chiller-cooling mode. A nonlinear optimization model for the PSAC system operation is developed and solved by the dynamic programming method to achieve global optimal results. According to the photovoltaic generation and building cooling load predictions

several case studies have been conducted for PSAC predictive control with different building cooling loads. The results show that when the photovoltaic generation prediction is considered

the cooling load for forming ice at night is shifted to the daytime

resulting in a reduction of the system operation cost and improvements in the solar fraction

self-consumption ratio

and average COP. When the selling price of photovoltaic electricity decreases

the PSAC is able to utilize the photovoltaic electricity in a better manner and the system operation cost can be reduced further under the predictive control.