A hierarchical control strategy for reliable lane changes considering optimal path and lane-changing time point

Implementing reliable lane changes is crucial for reducing collisions and enhancing traffic safety. However, existing research lacks comprehensive investigation into the optimal path for maintaining driving quality, and little attention has been given to determining the appropriate lane changing time point. This paper addresses these gaps by presenting a novel hierarchical strategy. First, a synthesized safety distance for lane changing, which considers variable execution duration, is designed to reduce collision risk. Next, a hierarchy of optimization control strategies is proposed to obtain the optimal path. An upper neural network-fuzzy control algorithm is established to identify an appropriate lane-changing time point. Additionally, a lower neural network-improved firefly algorithm is formulated to optimize the preliminary safety path based on multiple driving criteria. Furthermore, the dynamics characteristics of the vehicle are incorporated into the model predictive control algorithm to ensure the vehicle follows the optimal path. Finally, the feasibility of the proposed hierarchical control strategy is validated through typical lane-changing scenarios conducted on the Carsim-Simulink platform. This study proposes a hierarchical optimization control strategy that includes a synthesized safety distance for lane changing, an upper neural network-fuzzy control algorithm to determine the lane changing timing, and a lower neural network-improved firefly algorithm to optimize the safety path based on multiple driving criteria.image