Nowadays, with the development of automotive industry, air pollution and energy shortage problems are becoming more and more serious. Developing hybrid electric vehicles is a possible solution to these problems. A hybrid electric vehicle has two power sources:electric motor and internal combustion engine. There is a large battery that works as an energy buffer. Thus, the working point of engine can be adjusted to the better. Because of the fact that city buses usually run on a crowded road, the engine tends to work with low efficiency. Therefore, it is suitable to use hybrid electric buses in cities.Energy management strategy is a core technology of hybrid electric vehicles. Currently, rule-based energy management strategies are commonly used in hybrid buses. These strategies are practical and easy to realize, but the benefits are not satisfying. Because buses run on a fixed route, the prediction of future power demand can be made. And it can lead to an optimization-based strategy which can potentially increase fuel economy.The following contents are included in the presented research:(1)Firstly, the velocity data along a certain bus route is collected with the help of global positioning systems (GPS). After analyzing the collected data, a speed prediction model and a Markov speed transfer model are established. The latter two energy management approach is based on these two models.(2)Design a real-time control strategy based on the speed prediction model. The strategy converts the origin problem into two sub-problems and solves the problems by greedy algorithm. The states and optimal control variables are updated along the bus route just like Model Predictive control. Then use the Advanced Vehicle Simulator (ADVISOR) in Matlab to verify the effect of the strategy.(3)Design an energy management strategy based on the Markov speed transfer model. Through the analysis of the related state variables, the optimization problem is formulated as a Markov Decision Process. Then, the problem is solved offline and the results under different initial states are saved in a table. During the simulation, the optimal decisions are made according to the table. Finally, the comparison between these two strategies is given.Simulation work in the Advanced Vehicle Simulator (ADVISOR) shows that the presented approach can reach better fuel economy than the electric assist control strategy. |