Autonomous recognition system and method for uncertain poses of robot

Abstract

The invention discloses an autonomous recognition system and method for uncertain poses of a robot, the autonomous recognition system comprises the robot and an environment marker, and the robot comprises a control and calculation unit, a motion unit, a visual sensor and a laser ranging sensor; the environment markers comprise a primary position marker and an accurate position marker, autonomous perception of the environment is achieved through equipment of the robot, then the current pose of the robot is recognized, the robot does not depend on support of an external intelligent equipment system, and the complexity and cost of the system are reduced; and through pose deviation compensation, the situation that the robot reaches the standard operation pose after multiple times of pose adjustment can be avoided, reasonable pose deviation can be allowed, and it is ensured that the robot reaches the working result consistent with the standard operation pose under the non-standard operation pose, so that the robot can work after entering an operation point at a time, multiple times of position adjustment is avoided, and the working efficiency is improved. The robot working efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of intelligent robots, in particular to an autonomous recognition system and method for uncertain poses of robots. Background technique [0002] At present, in industrial robots and life service robots, it is often necessary to travel to and from a specific location (working location) to complete some operations at a specific location. For example, warehouse robots need to go to a fixed location to pick up and store goods, and service robots need to go to a fixed location to charge. [0003] Robots tend to operate more accurately in standard positions and have a higher success rate. When the robot leaves a specific standard location and then returns, it is often difficult to accurately enter the original standard operating point, and there is often a positional deviation; the deviation of the position and posture of the actual work location relative to the standard operating point will affect the accuracy o...

AI Technical Summary

Problems solved by technology

When the robot leaves a specific standard location and then returns, it is often difficult to accurately enter the original standard operating point, and there is often a positional deviation; the position and posture deviation of the actually entered working location relative to the standard operating point will affect the effect of operating accuracy, and often requires the robot to do it many times. It is inefficient to frequently adjust one's own pose to achieve a relatively accurate target work location, so there is an urgent need for an autonomous recognition system and method for robots with uncertain poses to solve the above problems

Images