Utilization of ANN and ANFIS models to predict variable speed scroll compressor with vapor injection

The use of numerical models and experimental setups to evaluate various parameters of variable speed scroll compressors with vapor injection (VSSCV) seems to be time-consuming, expensive, and fairly complex for engineers; hence development of an intelligent predictive model that is quick, simple to use, robust, and accurate in this field of study is worthwhile and highly necessary for work. In this regard, the paper presents two intelligent modeling approaches using an Artificial Neural Network (ANN) and an Adaptive Neuro Fuzzy Inference System (ANFIS) for the first time to accurately calculate the suction, discharge and injection mass flow rates (m˙SUC, m˙DIS, and m˙INJ), compressor electrical power (W˙COMP), and refrigerant temperature at compressor discharge (TDIS) for a VSSCV. The comparison between the developed models via statistical criteria showed the higher precision of applying the ANFIS approach as a suitable model for the prediction of VSSCV parameters compared to the ANN one.