Search Results

An energy management strategy is needed to optimally allocate the driver's power demands to different power sources in the fuel cell hybrid vehicles. The driver's power demand is modelled as a Markov process in which the transition probabilities are estimated on the basis of the observed sample paths. The Markov Decision Process (MDP) theory is applied to design a stochastic energy management strategy for fuel cell hybrid vehicles. This obtained control strategy was then tested on a real time simulation platform of the fuel cell hybrid vehicles. In comparison to the other 3 strategies, the constant bus voltage strategy, the static optimization strategy and the dynamic programming strategy, simulations in the Beijing bus driving cycle demonstrate that the obtained stochastic energy management strategy can achieve better performance in fuel economy in the same demand of dynamic.

Battery is a vital part of a fuel cell hybrid vehicle, and also the most difficult part to model due to its nonlinearity. Therefore, This paper presents an integrated software-hardware solution to simulate the fuel cell vehicle power train more accurately based on battery-in-the-loop, with the aid of RT-LAB™. Moreover, the average modeling technique is used together with RT-LAB's distributed cluster technology to realize real-time simulation of the Field-Oriented Controlled induction motor drive, and the Boost DC/DC converter. As a result, a virtual test bed, which is very similar to actual power train, is set up. Finally, on this test bed some tests are performed to verify the existing battery model and soc estimation method, and to give more accurate fuel consumption results.