Fuzzy modeling and analysis for cold storage capacity of capsulated plate ice storage system

Based on the non-linear characteristics in the cold storage capacity of ice storage system,the adaptive neural network was applied to the fuzzy modeling for the initialization and real-time operation of capsulated plate ice storage device. The membership function of such influencing factors as coolant inlet temperature,refrigerant flow and ice storage time as well as the fuzzy inference rule were determined. For the ice-storage system to track the load condition,the initialization training and operation comparison for the model were conducted with subtractive clustering and gridding methods. The ice storage quantity may be more reasonably initialized through adjusting and controlling the parameters of influencing factors. Further,the modeling programme was used to validate the fuzzy modelling strategy of the ice thaw outlet temperature. The operation of the model is relatively stable. The fuzzy control performance optimization for the cold storage capacity of the ice storage device ensures the matching of the cold storage capacity with load and the real-time control of the output ice thaw outlet temperature. Thus,the optimization and economical operation of ice-storage air-conditioning system can be achieved.