A refrigerant cireuitry qptimization method using an ineger gadient descent algorithm for hea pump finned-tube evaporators was devebped .A zone-based evaponlor simulation model was built

and the entopy genention variations with different evaporator refrigenant circuits were invetigated. It was concluded that when the designed heat transfer capacity increases. refrigerant passes should be increased

and the split point should be moved forward to minimize evaporator entropy generation. Subsequently

experiments were conducted to validate the relationship between the lower evaporator entropy generation (theoretical results) and the higher COP (experimental results) of a specific heat pump system. The evaporator entropy generation based on simulation results proved an effective parameter for predicting system COP variation trends. Subsequently

evaporator entropy generation was selected as the cost function

and an integer gradient descent algorithm was applied to minimize the evaporator entropy generation. With the proposed optimization method

the optimum evaporator refrigerant circuits corresponding to the heat transfer capacity in the range of 2000–6000 W were selected.