Adaptive variable step-size extrication strategy of the intelligent vehicle for the sudden concave obstacle in the garage scenario

Abstract

Aiming at the problem of the intelligent vehicle encountering the sudden concave obstacle and being unable to extricate under the garage scenario, this paper proposes an adaptive variable step-size extrication strategy based on the improved Hybrid A* algorithm, which guarantees path smoothing and enables effective extrication. Firstly, the collision constraints are established based on the collision detection model, where the double-circle model is employed, such that the proposed method is computationally less intensive while having a low redundancy space. Secondly, an improved Hybrid A* algorithm is proposed, which adds the node position penalty term based on the vehicle body coordinate system in the extension process, by which, the smoothness of paths can be guaranteed and the number of extended nodes can be reduced greatly. Finally, the relevant sizes between the vehicle and the concave obstacle, and the detection distance of the detection equipment are used as reference information to present an adaptive variable step-size extrication strategy. Simulation experiments of the improved Hybrid A* algorithm and adaptive variable step-size strategy are performed on Matlab software, where the vertical garage and the diagonal 45° garage are used as two different scenarios, and the results verify the feasibility and effectiveness of the strategy.