Variable-step-length A∗ algorithm for path planning of mobile robot

Abstract

The path planned by the traditional A∗ algorithm is not the shortest and there are too many steps because of the one grid step length. To overcome the drawbacks, an improved A∗ algorithm based on variable-step-length is presented for path planning of mobile robots. The environment of the mobile robot is modeled by the grid method. Considering that the reachable region of the mobile robot may be larger than one grid in each searching step, a variable-step-length scheme is proposed. The path planned in each cycle can be selected from the lines pointing to every grid in the reachable range surrounding the current grid. For the sake that the line distance between two points is the shortest, the planning indices such as the path distance and the step number are all better than those of the polygonal line planned by the traditional A∗ algorithm. Base on the variable-step-length, the searching direction and the improved cost function are designed for the optimal step. The corresponding obstacle avoidance algorithm is also discussed. The proposed variable-step-length A∗ algorithm make the mobile robot to choose an appropriate step according to the current environment and find a shorter and less steps path. The simulation results show the improved algorithm is more efficient for the path planning of the mobile robot.