Two-layer eco-driving approach of connected hybrid electric vehicles by convex optimization via signalized intersections

Abstract

Developing a sophisticated energy-management-centered eco-driving method can significantly boost the driving economy of vehicles. However, current optimization methods for hybrid electric vehicles (HEV) in multi-traffic-light scenarios still have room to improve energy-saving optimality and computational efficiency. Hence, this paper proposes a two-layer convex approach for the eco-driving of connected HEVs at signalized intersections. In the upper layer, a convex motor-power model is built, and a position constraint within the green-light time window is decided using traffic light information and signal phase-and-timing data. Then, a convex velocity-planning problem to minimize motor energy consumption is formulated and efficiently solved. In the lower layer, the engine's optimal operating line and unified battery-power constraints are introduced, and a series of convexification steps are performed. This enables the establishment of a convex-optimization energy management problem for minimizing fuel consumption, facilitating fast solution. Results show that the proposed method can effectively manage multi-signal scenarios, allowing vehicles to pass through green lights and avoid red-light waits. Regarding motor energy and fuel consumption, it achieves near-optimal results, with a deviation of less than 2 % from the global optimum. The optimization takes only about 1 s (around 1/24000–1/6 to comparative methods’), indicating high computational efficiency.