Improved Energy Management with Vehicle Speed and Weight Recognition for Hybrid Commercial Vehicles

The driving conditions of commercial logistics vehicles have the characteristics of combined urban and suburban roads with relatively fixed mileage and cargo load alteration, which affect the vehicular fuel economy. To this end, an adaptive equivalent consumption minimization strategy (A-ECMS) with vehicle speed and weight recognition is proposed to improve the fuel economy for a range-extender electric van for logistics in this work. The driving conditions are divided into nine representative groups with different vehicle speed and weight statuses, and the driving patterns are recognized with the use of the bagged trees algorithm through vehicle simulations. In order to generate the reference SOC near the optimal values, the optimal SOC trajectories under the typical driving cycles with different loads are solved by the shooting method and the optimal slopes for these nine patterns are obtained. When applying the developed strategy on the road, the driving pattern is timely identified and updated every 5 km by the model using the vehicle speed and driving power data in the past 500 seconds. Based on the recognized results, the reference SOC is then planned by selecting the corresponding pattern’s optimal SOC slope. Finally, a proportional control based on the SOC feedback is employed to track the reference SOC trajectory and optimize the fuel economy. The experimental and simulated results indicate that the proposed strategy has a fuel-saving ranging from 5.87% to 8.25%, with the highest value under the off-load cycle. The results also show that the impact of speed recognition on fuel consumption is more significant than that of load recognition.