There is a problem in that the periodic maintenance cost is high in a photovoltaic system and cannot make full use of the real-time state of the system to dynamically adjust the maintenance strategy. Thus a new dynamically combined maintenance strategy for a photovoltaic system based on associative set and availability is proposed by considering the influence of equipment operational status and weather factors. First, the associative set between the equipment from the functional correlation is established and used to make maintenance decisions. Then the status of equipment and the associative set is predicted through an improved Markov model and their availability is calculated. Secondly, in the stage of equipment maintenance decision-making, the lowest maintenance cost and the highest availability are the goals. The maintenance time and mode of equipment is determined by the mode factor and the equipment status. In the optimization stage of associative set maintenance strategy, a joint maintenance decision set based on the economic and structural correlation of equipments is established and the best maintenance strategy for the system is determined. Finally, a photovoltaic power station is taken as an example to verify the correctness and effectiveness of the proposed method. The results show that this method can effectively save on the total cost of system maintenance and improve the availability of the system. This work is supported by the National Natural Science Foundation of China (No. 51767017).