The invention discloses a low-computing-power dynamic
obstacle avoidance method based on single 2D
laser radar data. The method comprises the following steps: 1, collecting and preprocessing
laser radar data; 2, modeling an obstacle in the environment, converting the
laser point cloud data into the position and shape of the obstacle, performing local path planning with the minimum
path width through an incremental local path planning
algorithm, dynamically adjusting the local path according to the real-time laser
point cloud data, and obtaining the position and shape of the obstacle; only calculating a path through which the equipment can pass through the obstacle and reach the end point at the current moment each time; and 3, defining a cost function considering the
path length, the distance from the path point to the obstacle, the
turning angle and the equipment size, obtaining the obstacle information according to the real-time laser
point cloud data, calculating the
shortest distance between the current path and the obstacle, and if the
shortest distance is smaller than a safety threshold, adjusting the
weight coefficient of the cost function, and locally optimizing the path. According to the method, the path planning performance under a low-computing-power platform is improved, and equipment can efficiently and dynamically avoid obstacles in real time.