Open Issues in Supervisory Control of Hybrid Electric Vehicles: A Unified Approach Using Optimal Control Methods.

Abstract : Energy management of hybrid propulsion systems is considered, presenting new issues that extend the energy management role beyond the standard torque splitting to maximize system efficiency. The new issues include additional optimization criteria, constraints and relevant dynamics to deal with. New optimization criteria in addition the sole fuel consumption minimization include engine-out pollutant emissions and battery aging. State constraints are modified to account for plug-in hybrid vehicles. Moreover, specific supervisory control problems are recognized to need additional state variables. The latter comprise: engine and catalyst temperature to deal with engine warm-up effects on fuel consumption and after-catalyst emissions; thermal dynamics of heat recovery systems (Rankine or Thermo-Electric Generators, TEGs); and battery temperature, which influences battery performance and aging. It is shown that all these control problems can be treated in an unified fashion by extending the well-known ECMS (Equivalent Consumption Minimization Strategy), which is an implementation of Pontryagin Minimum Principle (PMP) as formulated by optimal control theory. Extended definitions of the Hamiltonian function and Lagrange multipliers are introduced. Optimization runs performed off line are reported. Results show the benefits of the proposed unified approach and enlighten some first online implementation issues.