Modeling of a hybrid ejector air conditioning system using artificial neural networks

In order to predict the performance of a hybrid ejector air conditioning system, neural network is chosen to model the proposed platform. First, three different types of neural networks, namely multi-layer perceptron (MLP), radial basis function (RBF) and support vector machine (SVM) are applied to model the component of a hybrid ejector air conditioning system. The MLP outperforms other two networks in this research and therefore it is selected to model the whole system. Since there is no formal criterion about input selection so far, a date-mining algorithm, boosting tree, is employed before system modeling to search the most significant parameters among the 19 input variables and the five most influential parameters of them are selected to be the final input of the system model. And the result shows a good agreement between predicted and measured value which indicates the excellent ability of MLP.