Improving the measuring length accuracy of articulated arm coordinate measuring machine using artificial neural network

作者：Zhao, Huining; Yu, Liandong^*; Cao, Jiaming; Jia, Huakun; Jiang, Yizhou

来源：Review of Scientific Instruments, 2020, 91(12): 125106.

DOI：10.1063/5.0028365

• MIND EVOLUTIONARY ALGORITHM
• SYSTEM-IDENTIFICATION
• CALIBRATION
• MODEL

A new method for further improving the measuring length accuracy of the articulated arm coordinate measuring machine (AACMM) is proposed. The detailed procedure of the proposed method involves kinematic error calibration with the Levenberg-Marquardt algorithm and then non-kinematic error (such as link deflection, thermal errors, and error motions of the rotation shaft) compensation with a back-propagation neural network optimized by the mind evolutionary algorithm. In order to verify the effectiveness and correctness of the proposed method, the simulation and experiment were carried out on an AACMM. The simulated and experimental results demonstrate that the measuring length accuracy of the AACMM is improved significantly after kinematic error calibration and non-kinematic error compensation, confirming the effectiveness and correctness of the proposed method.

• 单位
  合肥工业大学