Application of AFNNC to fresh air-conditioning system

A fresh air system with solar energy and phase-change thermal storage is introduced in this paper. It can not only supply fresh air to the room, but also supply parts of the heat to the room as a type of auxiliary heating to some extent. For the mathematical model of the fresh air system was difficult to accurately establish, an AFNNC (adaptive neuro-fuzzy inference system controller) was presented. It was used for parameter identification of input variables membership function and output variables applying for BP neural network back-propagation algorithm. Experimental results show that: The algorithm combining neural networks and fuzzy inference modeling and parameter identification methods can largely close to the practical data that solar energy - phase change thermal storage characteristics of the fresh air system, the error is almost zero. It finally proposed a new control method for the fresh air system which provides another idea for better energy saving.