Computationally efficient velocity and power split control of hybrid electric vehicles

An assessment study is presented for online optimal control of hybrid electric vehicles (HEV), based on the Pontryagin's Maximum Principle. A parallel HEV is used as an example, where states in the optimal control problem are the electric energy storage, kinetic energy and travel time. The solution quality and computational effort of the proposed method are assessed by comparing the results with a benchmark method that yields optimal performance. It is illustrated that the proposed method yields a good solution by solving the optimal control problem over 50 times faster than the benchmark method. Additionally, the novel approach brings up to 16 % decline in fuel consumption compared to a simple velocity trajectory which a human driver would choose.