Artificial neural netwroks-based modeling for electrical actuated automotive cooling systems

In this study, artificial neural network-based models for electrical actuated automotive cooling systems are developed and analyzed. The proposed models are constructed to represent the dynamical behavior of advanced automotive cooling systems. Three types of artificial neural networks (ANNs) are constructed; MPL, Elman, and NARX networks where experimental data are utilized in the development process of the models. The proposed ANN models are developed and tested in simulation to approximate the performance of the actual system. Statistical results are demonstrated to show the effectiveness of the proposed ANNs.