Namwook Kim1, Jongryeol Jeong, Chunhua Zheng
Abstract
Optimal control ideas based on Pontryagin’s Minimum Principle (PMP) have become mature techniques for maximizing the fuel efficiency of Hybrid Electric Vehicles (HEVs) and Plug-in Hybrid Electric Vehicles (PHEVs). The outstanding performance of this control concept has already been verified in many studies, in which the PMP-based control produces optimal solutions that are very close to the global optimal solution obtained by Dynamic Programming (DP). However, the drawback of the control concept is that the PMP-based control will not guarantee optimality if no information about the future driving condition is given. This is not just a drawback of the PMP-based control, but it is an unavoidable limitation in most optimal control concepts. Therefore, previous studies have been focused on finding an optimal costate when the future driving conditions are given or predicted prior to driving. In this study, a methodology that analyzes the past driving pattern and updates the control parameters is proposed by assuming that vehicles are operated under repeated driving conditions. A control parameter, or a costate in the PMP-based control, can be estimated from two parameters that characterize the driving conditions, and the correlation between the costates and the energy consumption patterns is used to update the control parameter. Based on this control concept, the final State of Charge (SOC) at the end of each drive gets gradually closer to the desired value as the driving cycle is repeated. The methodology can be used for vehicles operated under repeated driving patterns, such as commuting buses, parcel delivery vehicles, or refuse collection trucks.