移动机器人导航规划的双向平滑A-star算法

Aiming at solving the problems in existing mobile robot navigation, such as lower-precision localization of the MCL algorithm, poor real-time ability of the traditional A-star algorithm and excessive polylines of the planned path, this paper proposes a novel A-staralgorithm- based mobile smooth navigation strategy. Firstly, the multi-sensor Monte Carlo localization (MS-MCL) algorithm is proposed for the pose estimation. Secondly, by improving the A-star algorithm to the bi-directional pattern and using the cost function iteratively, the algorithm has better real-time performance. Finally, Bézier curve is used to optimize the planned path, which solves the problem that the algorithm cannot be applied in practice because of excessive polylines and tiny turning angles. In comparison with the traditional A-star trajectory planning algorithm, simulation experiments fully proved that the algorithm designed in this paper reduces the path length and running time by 12.1% and 37.2%, respectively, and the average safety distance and the path smoothness are increased by 35.2% and 69.9%, respectively. Meanwhile, the results of the practical experiment verify the correctness and feasibility of this navigation planning algorithm under the guaranteed localization accuracy.