The vast majority of industrial robots are model-based Control, where there is a Model, there will be errors, which needs to be calibrated. Industrial robot calibration is a process that integrates modeling, measurement, robot parameter identification and error compensation.
In order to calibrate the robot, we must first understand the possible factors that cause the error of the robot.
There are many factors that cause robot error, which can be summed up into two categories. One is kinematic correlation, which mainly includes machining error, mechanical tolerance/assembly error, zero point error, reducer return error, deceleration ratio error, calibration error, etc. The second is kinematic correlation, which mainly includes mass/center of gravity, inertia tensor, friction force, joint flexibility, linkage flexibility, etc.
The practice found that the robot base without calibration usually has an error of 15~30mm, the TCP center end has an error of 5~10mm, and the whole robot system has an error of 5~10mm. After calibration, the error range of the robot can be greatly reduced.
· If the robot is not calibrated, the robot cannot share the program, and the accuracy is very low and unstable;
· Calibration can improve the robot's ability to deal with environmental uncertainties;
· The main reason for the low accuracy of the robot is that the design parameters of the robot are different from the actual parameters, and the robot calibration is a measure to improve the accuracy of the robot, which can often increase the accuracy by several orders of magnitude;
· In the process of robot development, sufficient accurate data must be obtained to analyze and evaluate the machine static and dynamic, and these data are obtained by a complete set of calibration system.
