Co-Optimization of Velocity Planning and Energy Management for Intelligent Plug-in Hybrid Electric Vehicles Based on Adaptive Dynamic Programming

Intelligent plug-in hybrid electric vehicles (IPHEVs), with advanced environmental sensing and integrated communication capabilities, present tremendous autonomy in drive decision-making and high efficiency in energy economy. In this context, the key control of IPHEVs gradually evolves from energy management among the key components of powertrain to the co-operative velocity planning and energy management, which, nonetheless, are strongly coupled. To tackle it, an intelligent strategy based on adaptive dynamic programming (ADP) is explored to online co-optimize the longitudinal traction and energy management of IPHEV, especially in vehicle-following scenarios. The imbedded cooperative critic and actor networks are constructed to endow the ADP with satisfactory learning capability and calculation speed. The instantaneous reward integrating the nonlinear characteristics of IPHEV powertrain is designed for online learning of critic and actor networks. Substantial simulations demonstrate that the developed ADP is computationally efficient to ensure the driving comfort, safety, and fuel economy of the studied IPHEV when simultaneously planning the vehicle speed and allocating the power within the powertrain. To be specific, the developed strategy can reach 98.01% and 94.22% energy optimality of dynamic programming under urban and mixed driving conditions.