Modelling and predicting energy consumption of a range extender fuel cell hybrid vehicle

Energy consumption is an important economical index of a fuel cell hybrid vehicle (FCHV). To analyse the energy consumption of a range extender FCHV and reduce the cost of experiments, this study developed a nonlinear regression model of the powertrain of the vehicle to predict the current and voltage on the DC bus, which were used in the investigation of energy consumption, by using the intelligent algorithms including Back Propagation neural network (BP), Genetic Algorithm-Back Propagation neural network (GABP) and least square support vector machine (LSSVM). The model based on the LSSVM achieves the best predicted performance and can consider the nonlinear characteristics of the powertrain quite well. A case study was discussed by applying the obtained model and integrated with a hierarchical energy management strategy (HEMS). The specific results of energy consumption showed that it is feasible to use the predicted data of the obtained model in the analysis of the energy consumption of the FCHV.