Modeling, implementation and experimental verification of eco-driving on a battery-electric heavy-duty vehicle

In this paper, an Eco-Driving Assistance System (EDAS) has been implemented on a fully electric heavy-duty vehicle and its performance has been validated using real-world experiments. The objective of the EDAS is to provide the driver with a recommendation on the vehicle’s optimal speed trajectory that minimizes its energy consumption over the entire trip. This requires solving a receding horizon optimal control problem, which, in this case, consists of a convex optimization problem and can be solved as a second-order cone program. Simulations were used to explore different prediction horizon lengths and move-blocking strategies of the underlying receding horizon optimal control problem, aiming to strike a balance between numerical complexity and energy savings. Finally, the method is implemented on an electric heavy-duty vehicle where an augmented speedometer is presented to the driver. Multiple tests with and without an EDAS have been performed, which resulted in a reduction of 6.5 %–12 % in energy consumption compared to when the vehicle was driven without the EDAS active. • Eco-driving coaches drivers to optimize speed and driving behavior for improved energy efficiency. • The proposed eco-driving problem leverages convex optimization to enable real-time on-vehicle implementation. • On-road tests are conducted using a real vehicle in real traffic conditions, with a human driver receiving instructions via a human-machine interface. • Experiments show a 6.5 %–12 % reduction in vehicle energy usage compared to that of a non-coached human driver.