滑动转向机器人路径规划算法研究

iming at the problem that the skid-steered robots will slip when turning, which makes it difficult to complete the navigation task smoothly and efficiently, a path planning algorithm for the skid-steered robots is presented. In the global planning, the improved A-star algorithm is used to improve the problem of unsmooth paths planned by the traditional A-star algorithm, and the gradient descent method is used for smoothing, and the smoothing effect of different smoothing parameters is compared and tested. The parameters suitable for skid-steered robots are presented. At the same time, in the local planning, the dynamic arc fitting algorithm is improved, and the dynamic look-ahead distance algorithm is added to make the arc radius larger and smoother, which is more suitable for the structural characteristics of the skid-steered robots, and makes the local path planning can better fit the global reference path. After simulation experiments, the improved dynamic arc fitting algorithm can reduce the maximum fitting error by 21.85cm, which proves its effectiveness.