A novel neural network aided fuzzy logic controller for a variable speed (VS) direct expansion (DX) air conditioning (A/C) system

A novel artificial neural network (ANN) aided fuzzy logic controller for simultaneous control of indoor air temperature and humidity using a variable speed (VS) direct expansion (DX) air conditioning (A/C) system, through combining the complementary merits of fuzzy logic controllers and ANN modeling was developed and is reported in this paper. A novel control principle was proposed to decouple the temperature and humidity control loops by introducing two interim variables of sensible and latent output cooling capacity of the DX A/C system. A fuzzy logic system was redesigned to simplify both its calculation and structure by using weights of linguistic variables. To enable the ANN model developed to be functional at the normal operational range of indoor air parameters, previously reported inherent operating characteristics of a VS DX A/C system were used for training and testing the ANN models. The novel controller so developed was tested using an experimental VS DX A/C system. Both the command following tests and disturbance rejection tests showed that the air dry-bulb and wet-bulb temperatures were properly controlled by the controller developed with satisfactory control performances in terms of control accuracy and sensitivity.