Energy Management Design in Hybrid Electric Vehicles: A Novel Optimality and Stability Framework

摘要

This paper addresses the problem of finding a closed-form optimal solution for the energy management problem in charge-sustaining hybrid electric vehicles (HEVs), and proposes, for the first time, a generalized stability and optimality framework for this type of problem. The energy management problem, which by its very nature is a finite-time horizon control problem, is reformulated as a nonlinear-nonquadratic infinite-time optimization problem, leading to a family of state-feedback control laws that provide optimality with respect to an infinite time horizon performance objective, while guaranteeing asymptotic stability. The stability problem in charge-sustaining HEVs is formulated to allow the design of analytical solutions using a Lyapunov-based argument. The proposed control law is implemented on a pre-transmission parallel hybrid heavy-duty vehicle and the performance of the closed-loop system is shown in simulation and compared with the benchmark solution provided by Pontryagin's minimum principle (PMP) and the real-time adaptive controller adaptive-PMP. Results show low sensitivity to the control parameter, low-calibration effort, and reduction of computational effort, while maintaining close-to-the-optimum performance. Hardware-in-the-loop simulations were conducted to validate and verify the new strategy in a real-time simulation setup.