Abstract
In this work, a new method id described and applied to find the global solution of the energy management problem, which is defined as finding the minimum overall fuel consumption for a known drive cycle by managing the energy flow through the engine, the motor/generator and the battery feasibly. Finding the optimal solution to this problem while meeting the hard constraints such as meeting the expected level of the battery energy at the end of the journey (end-charge SOC) is a challenging task. However, a hill-climbing energy management strategy (HCEMS) is demonstrated that starts with a feasible initial solution and evolves towards the final solution always in the direction of solution improvement. The objective function consists of the actual fuel consumption, a penalty to reduce the final deviation of the battery energy and a penalty to reduce frequent stopping and starting of the engine. The HCEMS is compared with two existing energy management techniques dynamic programming (DP) and equivalent consumption minimization strategy (ECMS). Fuel consumption solutions of the HCEMS are nearly as good or equal to DP. HCEMS requires very low computational cost compared to DP. Furthermore, HCEMS always performs better than ECMS in terms of fuel consumption and achieving the end-charge sustenance. The solutions are demonstrated on a E segment vehicle configured for close to optimum fuel consumption with a 50 kW high efficiency engine and 30 kW motor that gives 13s 0-100km/h acceleration time. This powertrain is reconfigured with an ultra-lean burn super-charged engine with 120 kW and very high efficiency, giving an acceptable 7.5s 0-100km/h. These vehicles have a modelled NEDC CO2 of below 90 g/km even when the electric motor in the large engined powertrain has the same 30kW motor.
KEYWORDS – HEV energy management strategy, hill-climbing heuristics, dynamic programming, advanced gasoline engine, very low fuel use.