Optimal control of a parallel hybrid vehicle with a traffic preview

摘要

In this work, a short preview of traffic provided by on-board sensors is used to minimize fuel consumption in a parallel hybrid vehicle capable of controlling both the vehicle velocity and the torque-split ratio. A two-stage version of Pontryagin's minimum principle was used, but some of the standard simplifying assumptions in reduced-order modelling of hybrid vehicles were avoided to ensure that the fuel economy is not unduly compromised. The proposed controller is implemented in a receding-horizon fashion to incorporate updated telemetry information at each sampling instant. The effectiveness of the proposed method is demonstrated by simulations on two hybrid vehicle models containing higher-order dynamics and non-linearities. It is shown that the controller maintains the engine operation within the high-engine-efficiency region, and successive improvements in the fuel economy are achieved as the traffic preview length is increased.