Path planning method for robot leaded by global planner

Abstract

The invention relates to a path planning method for a robot leaded by a global planner. The method comprises the following steps of SLAM, construction of the three-dimensional robot, robot locating, construction of a local raster map, setting of an end point, global path planning, local path planning, updating of global and local raster maps and updating of robot states. Evaluation for global pathplanning is cancelled, although there is the global planner, actually, a path generated by the global planner is not evaluated, an existing local path planner and an existing local path evaluator aremainly improved, and traditional evaluation factors are revised as the number of barriers, the straight-line distance between the path and the nearest barrier, the path curvature and the like, wherein distances between the barriers and the robot are less than some value (depending on the actual environment). According to the method, data of a laser radar and ultrasonic waves is combined, and theproblem that a robot based on the laser radar cannot avoid transparent barriers is solved.

Description

technical field [0001] The invention relates to the field of robots, specifically, a two-dimensional navigation algorithm with a global planner as the main and a local planner as the auxiliary. Background technique [0002] Path planning refers to the optimal path optimization problem of robots. It has a rich application background, including sweeping robots and unmanned driving. The current common navigation algorithms are mainly based on local planners, supplemented by global planners, which are suitable for scenarios with high frequency of environmental changes, such as DWA dynamic window algorithm. These navigation algorithms need to evaluate the global path, which consumes unnecessary computing resources in indoor environments where the frequency of environmental changes is low. At the same time, the local path planners of these navigation algorithms generate too many paths, which are not suitable for indoor environments with low frequency of environmental changes. The...

AI Technical Summary

Problems solved by technology

These navigation algorithms need to evaluate the global path, which consumes unnecessary computing resources in indoor environments where the frequency of environmental changes is low

At the same time, the local path planners of these navigation algorithms generate too many paths, which are not suitable for indoor environments with low frequency of environmental changes. The evaluation factors of the local path evaluators are also too complicated, which takes up a lot of computing resources.

Images