Robot Calibration via Pose Constraint and Force Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot kinematic calibration methods rely on expensive and cumbersome external precision measuring equipment, which are not suitable for regular field calibration due to high costs, maintenance requirements, and limited portability. These methods also struggle to provide full-pose error information and compensate for non-geometric errors.
Innovation Solution
A robot calibration method based on pose constraint and force sensing, which utilizes a robot calibration device consisting of an end calibration device with a force sensor and calibration spheres, and a geometric constraint device with V-shaped grooves. The method involves establishing kinematic, geometric, and non-geometric error models, collecting data through drag teaching, and compensating parameter errors in the robot's controller.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If external precision measuring equipment (laser tracker, coordinate measuring machine) is used for robot calibration, then measurement precision is improved, but device complexity and cost increase significantly
Solution Approach 1:
The calibration device enables the robot to perform self-calibration through drag teaching operation. The robot end effector automatically interacts with the calibration device's geometric features (V-shaped grooves, cylindrical surfaces) to collect pose data, eliminating the need for complex external measuring equipment and manual operation.
Solution Approach 2:
A calibration device serving as an intermediary is introduced between the robot and the calibration process. This device contains precise geometric features (V-shaped grooves, cylindrical surfaces, spherical surfaces) that the robot end effector can detect and use to establish pose constraints, replacing complex external measuring equipment.
2Device complexity
If traditional position-based calibration devices are used, then device complexity is reduced, but measurement precision and error information completeness deteriorate
Solution Approach 1:
The calibration device merges multiple geometric constraint features (V-shaped grooves for position constraint, cylindrical surfaces and spherical surfaces for pose constraint) into a single integrated structure. This allows the simple device to provide comprehensive pose information (position and orientation) that traditional position-only devices cannot deliver.
3Device complexity
If geometric constraint devices without force sensing are used, then device complexity is reduced, but ability to detect non-geometric errors deteriorates
Solution Approach 1:
The calibration device achieves multi-functionality by combining geometric constraint features (for geometric error detection) with force sensing capabilities (for non-geometric error detection). This allows a single device to comprehensively detect both geometric and non-geometric errors without requiring separate specialized equipment.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method provides comprehensive all-pose error information and effectively estimates and compensates for non-geometric errors, offering a cost-effective and portable solution for robot calibration that surpasses traditional external calibration devices and most self-calibration devices.
Implementation Method 1
a force sensor, a connecting seat, and a plurality of calibration spheres. The connecting seat is connected to the force sensor
Data Source
AI summary
A robot calibration method based on pose constraint and force sensing includes the following steps: establishing a kinematic model, a geometric error model, and a non-geometric error model; installing an end calibration device to an end of a robot, and installing to the geometric constraint device into a working space of the robot; dragging the robot, constraining various calibration spheres of the end calibration device into various V-shaped grooves on the geometric constraint device to achieve pose constraint; then dragging the calibration spheres to V-shaped grooves on different surfaces, and calibrating a geometric parameter error of the robot using a deviation between a nominal end pose measured twice and an actual value; reading an end force by a force sensor to calibrate the non-geometric error model; identifying kinematic model parameters of a corresponding robot; and compensating an identified kinematic model parameter error to a controller of the robot.