Dimensionlessness and threshold denoising in multi-joint arm calibration model with distance constraints

Considering that the matrix of multi-joint arm calibration model with distance constraints, in which there were different orders of column magnitude, may cause ill-posed problem and the error in the process of calibration may affect the accuracy of the result, a method was proposed to modify the calibration matrix. Dimensionlessness and threshold denoising algorithms were used in the calibration matrix´s calculation. Recalibration and iteration were used to calculate the calibration matrix after dimensionlessness and threshold denoising. Experimental study was carried out on a 6-joint arm with a standard ruler, of which the distance was 711.199mm. The experiments showed that the method can protect the calibration equation from ill-posed problem; the standard deviation of length measurement was 0.103mm after dimensionlessness and threshold denoising, which were effective and suitable for the requirement of measurement.