Uncertain route, destination, and traffic predictions in energy management for hybrid, plug-in, and fuel-cell vehicles

Abstract

This paper incorporates uncertain future route predictions, destinations, and charging locations with associated speed and grade profiles into the energy management control of alternative powertrains like hybrid, plug-in, electric, and fuel cell vehicles. The method allows the combination of other sources of uncertain information like markov driver models, historic speed information, and real-time traffic predictions. This flexibility allows the consideration of a variety of information cases like uncertain traffic/speed and route information, multiple possible destinations, stopping points, and charging locations, simple range estimates to the destination, and no future knowledge at all. The model can be used with any vehicle type and stochastic control method, and is suitable for real-time calculations either on the vehicle or a server. Two techniques are also presented to reduce the computational complexity of the problem. This approach is demonstrated on a simulated trip with two possible destinations using the stochastic dynamic programming algorithm.