Experimental investigation on variable ejector in multi-evaporator refrigeration system using extreme learning machine

Abstract

In this report, the topics on multi-evaporator refrigeration system and extreme learning machine were extensively elaborated. Experiments were performed on the variable ejector in multi-evaporator refrigeration system so as to understand its operation further by expanding its system database. With the newly acquired data obtained from the experiments, the results show that its area distribution does increase. Subsequent tests were carried out to evaluate the system pressure recovery ratio with the addition of new data. The results were able to validate that the system’s PRR does improve as the cross-validation and test errors were reduced. An increase of 0.914% of PRR model accuracy was obtained. Thus, the conducted investigation supports previous researches and claims that the variable ejector is indeed a promising application in the multi-evaporator refrigeration system.