Performance prediction of a direct evaporative cooling air conditioner using neural network method

The mathematical model of direct evaporative cooling air conditioner is difficult to solve for the irregularity of pad surface. Recurring to the good nonlinear mapped performance of neural network and the experimental data of a direct evaporative cooling air conditioner with GLASdek pad, establishes a three layer back propagation neural network to predict the air handling performance of the conditioner under various working conditions. The predicting result proves consistent with the experimental data.