A PHEV Power Management Cyber-Physical System for On-Road Applications

Abstract

Power management of plug-in hybrid electric vehicles (PHEVs) helps achieve high energy efficiency and low energy consumption cost. Existing power management methods are classified into online and offline ones. Online methods are usually designed with preset power-balancing strategies to adapt to runtime conditions but without utilizing route information or historical driving cycle information. In offline methods, power management strategies are derived from historical driving cycles and not optimal for actual driving routes. Facilitated by vehicular network technologies, on-road PHEVs can easily upload their driving data through on-board communication devices, e.g., smartphones, to remote servers. With information sharing, individual vehicle speed prediction can be realized and used in the on-board power management system. This paper proposes an on-road PHEV power management cyber-physical system with two-level hierarchical strategies to minimize the fuel consumption in a trip. The high-level management strategies allocate battery energy budgets to all remaining roads, which are generated online by solving stochastic optimization problems. The low-level powertrain policies are solved offline from historical driving cycles. While a PHEV is driving, optimal real-time power decisions are made based on battery budgets, low-level power policies, and the PHEV's driving states. Simulation results show that the proposed method outperforms other methods significantly in terms of fuel savings.