Obstacle avoidance method and system for unmanned cleaning vehicle in dynamic environment

Abstract

The invention discloses an obstacle avoidance method and system for an unmanned cleaning vehicle in a dynamic environment, and the method comprises the following steps: building a factory scene map through a laser radar and a binocular camera which are disposed below the unmanned cleaning vehicle, and carrying out the layering of the map; then dividing the map into an original map layer and a comparison map layer, comparing and matching the comparison map layer and the original map layer, and marking distinguishing articles in the original map layer; classifying obstacles in the map, and further identifying pedestrian obstacles; and searching an optimal path by adopting a dwa algorithm.

Description

technical field [0001] The invention relates to the technical field of unmanned cleaning vehicles, in particular to an obstacle avoidance method and system for unmanned cleaning vehicles in a dynamic environment. Background technique [0002] Compared with manual cleaning, unmanned cleaning vehicles have been gradually applied due to their outstanding cleaning efficiency and economic cost. Due to the large work scene and complex obstacle information, unexpected obstacles will inevitably appear. The obstacle avoidance of unmanned cleaning vehicles is the first priority. Factors to be considered. At present, in terms of obstacle avoidance, most unmanned cleaning vehicle R&D and manufacturers use two-dimensional / three-dimensional laser radar or visual cameras to detect obstacles, and use slam technology or multi-sensor fusion. When there is an obstacle, control the vehicle to bypass the obstacle. However, the visual method is rich in perception information, but the amount of in...

AI Technical Summary

Problems solved by technology

[0002] Compared with manual cleaning, unmanned cleaning vehicles have been gradually applied due to their outstanding cleaning efficiency and economic cost. Due to the large work scene and complex obstacle information, unexpected obstacles will inevitably appear. The obstacle avoidance of unmanned cleaning vehicles is the first priority. Factors to be considered. At present, in terms of obstacle avoidance, most unmanned cleaning vehicle R&D and manufacturers use two-dimensional / three-dimensional laser radar or visual cameras to detect obstacles, and use slam technology or multi-sensor fusion. When there is an obstacle, control the vehicle to bypass the obstacle. However, the visual method is rich in perception information, but the amount of information is large, which has high requirements on the performance of the processor, and the cleaning speed cannot be too fast; although the radar obstacle avoidance method has a relatively large amount of data Small, the content of perception information for large scenes is limited, relying on software algorithms to calculate obstacle information, it may not be able to successfully bypass obstacles or the ability to circumvent obstacles is insufficient; in the method of sensor fusion, due to the simultaneous operation of multiple sensors, data processing is difficult. Most of the obstacle avoidance strategies are to avoid obstacles when encountering obstacles, and there is little classification of obstacle information, which also leads to bypassing when encountering workers or specific obstacles, while ignoring the area occupied by obstacles. deteriorating cleaning ability

Images