Location-based energy management optimization for hybrid hydraulic vehicles

Hybrid hydraulic vehicles are a promising approach towards improving the fuel efficiency of heavy vehicles such as garbage trucks and city buses. The combination of a conventional diesel engine with a hydraulic powertrain allows for regenerative braking which results in reduced fuel consumption, reduced emissions and less brake wear. Further improvements can be achieved by numerical optimization of the energy management strategy, i.e. the distribution of the desired torque among the two propulsion systems. However, for such strategies to be efficient, a short-term prediction of the driving profile becomes necessary, which is simply assumed to exist in most previous work. For the case of garbage trucks and city buses, the assumption of repeatedly driven routes is valid. Therefore, a system that iteratively learns driving profiles has been developed. The learned driving profiles are associated with a particular vehicle location. The results of prediction and optimization are validated in a simulation study based on a standard reference cycle.