DEVELOPING ANN MODEL TO EVALUATE COP OF VCR SYSTEM FOR SUCTION PRESSURE & TEMPERATURE, DELIVERY PRESSURE & TEMPERATURE AS INPUT PARAMETER
Rakesh Kumar Agrawal, G. K. Agrawal
ABSTRACT
Artificial neural network (ANN) network3 is developed to evaluate coefficient of performance of simple vapor compression system for various values of suction pressure, suction temperature, delivery pressure & delivery temperature as input parameter. 51 set of experimental set of parameters is used to train ANN, network3 and 12 set of experimental parameters is used to test ANN, network3. Value of experimental COP and that predicted by ANN network3 resemble close to each other with R2 = 0.9996346, RMSE = 0.111, COV=1.928 %. In developed ANN model network3, network type feed forward back propagation, Training function TRAINLM, Adaptation learning function LEARNGDM, performance function MES, no of layer 01, No of neuron 08 and Transfer function LOGSIG, with other training parameter has been used to successfully train ANN network3. It is concluded that ANN with developed network3 can be successfully applied for evaluation of coefficient of performance of simple vapor compression refrigeration (VCR) system & hence ANN may be very useful tool for performance analysis of refrigeration system.
[Full Text Article]
