A comprehensive control method for picking robots in shopping malls and supermarkets

Abstract

The invention discloses a comprehensive control method for a shopping mall and supermarket picking robot. The environmental image is loopback detected, and the Bayesian filtering algorithm is used to calculate the position coordinates of the shelf in the process of constructing the environmental map; 2) The path planning is carried out by combining the global optimal path planning and the local uniform path planning; using the fusion A* Algorithm and DWA algorithm for global optimal path planning, and local uniform path planning by means of uniform straight-line travel; 3) Real-time dynamic obstacle avoidance based on ultrasonic sensors, including non-picking obstacle avoidance and picking obstacle avoidance; 4) Based on single Target recognition for target cameras, including image annotation and dataset training. This method can realize the picking and delivery of robots in large scenes, which not only greatly reduces the demand for manpower, but also improves the adaptability of the scene.

Description

technical field [0001] The invention relates to the field of robot automatic control, in particular to a comprehensive control method for a shopping mall and supermarket picking robot. Background technique [0002] With the gradual entry of service robots into people's lives, especially the rapid development of e-commerce in recent years, picking robots have emerged as the times require. The picking robot can generally automatically complete the positioning, identification and picking of goods according to the picking list, and transport the picked goods to the designated location, thereby replacing heavy manual labor and realizing unmanned goods picking operations. In order to ensure that the picking robot can successfully complete the aforementioned actions, a set of control methods to realize the coordinated work of each part of the picking robot is particularly important. [0003] The Chinese patent with application number 201811371601.8 discloses a robot-based logistic...

AI Technical Summary

Problems solved by technology

[0003] The Chinese patent application number 201811371601.8 discloses a robot-based logistics partition picking method, device, terminal, system, and storage medium. After the robot obtains the location information of the goods, it then obtains the location information mapped by the robot’s picking location. Information, and then plan the path of the robot. This method needs to count the position information of all goods in advance. However, due to the variety of goods placed on supermarket shelves and the frequency of change of product positions, it takes a lot of time to count the position information of goods each time, and consumes a lot of manpower. , so this method is not suitable for supermarkets in complex environments

[0004] The Chinese patent application number 201710295472.8 discloses a robot control system, including an obstacle detection module, an obstacle processing module, and a navigation control module. It detects whether the obstacle is moving through the background modeling method or the frame difference method, and the static obstacle will Affects path planning, the robot walks around obstacles; dynamic obstacles do not affect robot path planning, the robot slows down when approaching dynamic obstacles, and tries to walk through dynamic obstacles; the robot needs to wait for the obstacles to be automatically eliminated before walking, affecting picking Therefore, this method is not suitable for complex occasions with many obstacles such as supermarkets.

Images