A Baseline Framework Model for an Emission-free Fuel Cell Vehicle System employing Highway and Federal Driving Procedure

In this paper the performance of Neural Network (NN) based fuel cell powered electric vehicle model is analyzed for three different modified drive cycle patterns such as M-HWY, M-US06 and M-FTP based on which they are operated. The complexity involved in the conventional mathematical modeling of the fuel cell stack system is eradicated with the Neural Network modeling. The Multilayer feed forward Neural Network is used to predict the output voltage of the PEM fuel cell from a predefined vehicle drive cycle pattern. The optimum neural network is chosen by varying the training algorithms and the number of neurons in the hidden layer. The performance comparison in terms of error minimization values such as MSE, MAE and iteration value is also carried out to verify the reliability of the optimum neural network. The optimum neural network chosen is used to develop a neural network based fuel cell driven electric vehicle model that includes the modeling of fuel cell, DC-DC converter and vehicle dynamics. The simulation results obtained from the developed electric vehicle model are used to evaluate the vehicle performance in terms of hydrogen consumption, maximum distance coverage and power flow within the vehicle system. The comparison of the required vehicle power with the fuel cell (Neural Network) delivered power is carried out to validate the optimality of the proposed electric vehicle model.