In the article “ABC”, author Baoyonghua proposed a new method for calibrating kinematic parameters of robots, aiming to improve the positioning accuracy of robots. The author proposed an improved bat algorithm to identify errors in robot kinematic parameters and presented a low-cost measurement method that uses a 3D scanner to measure the end position of the robot.
The author believed that the positioning accuracy of robots has a great impact on their applications. By modifying the kinematic model of robots without changing their structure, the positioning accuracy can be improved. The author outlined the sources of robot errors, existing calibration methods, and their limitations. Existing measurement methods are costly and require a lot of specialized knowledge; identification algorithms have slow convergence speeds and low accuracy.
To ensure the completeness, continuity, and minimality of the robot position error model, the author analyzed its redundancy and removed redundant parameters from it.
To address deficiencies in existing identification algorithms, the author proposed an improved bat algorithm to identify errors in kinematic parameters. The author provided a detailed introduction to this algorithm’s mathematical model and improvements. Natural selection was used as a strategy to increase convergence speed, while adaptive parameter control was employed to coordinate global search and local search.