Research On Error Identification Of Manipulator Based On Binocular Vision Measurement

The precision of manipulator is an important index to measure the performance of manipulator.In order to improve the accuracy of the manipulator,it is necessary to identify and compensate the error of the manipulator.Due to the high cost and complex operation of the commonly used measuring equipment such as laser tracker in the measurement process,an error identification method of the manipulator based on binocular vision measurement is proposed in this paper.The results are as follows:1.Taking LR7-710 robotic arm as the research object,after the selection of binocular measurement related equipment is completed,the non-parallel binocular vision structure model is proposed for image acquisition.In the process of measurement,it involves the conversion of the measuring coordinate system and the coordinate system based on the manipulator.Aiming at the problem that the conversion relationship is difficult to determine precisely,the error model based on the distance error is used to identify the DH parameters,and the relevant measurement scheme is designed.2.Research camera calibration technology,mainly including the acquisition of internal and external parameters of monocular camera and the determination of the position and pose relationship between binocular camera.In order to improve the accuracy of camera calibration,an improved particle swarm optimization algorithm was proposed to optimize the calibration parameters.3.In order to improve the accuracy of image feature point detection,the denoising algorithm and image segmentation algorithm in the process of image preprocessing as well as the positioning effect of checkerboard and concentric circles for the end of the manipulator were studied.A checkerboard corner detection algorithm based on growth is used to obtain subpixel corner coordinates for checkerboard targets.For concentric circle characteristic target,an improved morphological method combined with Zernike moment is proposed for target center sub-pixel localization.4.Experimental verification is carried out according to the method proposed in this paper.Firstly,an improved particle swarm optimization algorithm was used to optimize camera calibration parameters.Then two experiments were carried out to verify that the accuracy of the binocular measurement meets the requirement of the manipulator error measurement.Finally,the error parameter identification and compensation of the LR7-710 manipulator were completed,and the effectiveness of the error compensation was verified by controlling the manipulator to move according to the cube trajectory.Through the above research,it is verified that the binocular vision measurement method combined with the error model based on distance error is effective in the identification of the error parameters of the manipulator,and the positioning accuracy of the manipulator is improved after compensation.