A Novel Quadratic Interpolated Beetle Antennae Search for Manipulator Calibration

Over the past decades, industrial manipulators play a vital role in in various fields, like aircraft manufacturing and automobile manufacturing. However, an industrial manipulator without calibration suffers from its low absolute positioning accuracy, which extensively restricts its application in high-precision intelligent manufacture. Recent manipulator calibration methods are developed to address this issue, while they frequently encounter long-tail convergence and low calibration accuracy. To address this thorny issue, this work proposes a novel manipulator calibration method incorporating an extended Kalman filter with a Quadratic Interpolated Beetle Antennae Search algorithm. This paper has three-fold ideas: a) proposing a new Quadratic Interpolated Beetle Antennae Search algorithm to deal with the issue of local optimum and low convergence rate in a Beetle Antennae Search algorithm; b) adopting an extended Kalman filter algorithm to suppress non-Gaussian noises and c) developing a new manipulator calibration method incorporating an extended Kalman filter with a Quadratic Interpolated Beetle Antennae Search algorithm to calibrating a manipulator. Extensively experimental results on an ABB IRB120 industrial manipulator demonstrate that the proposed method achieves much higher calibration accuracy than several state-of-the-art calibration methods.