A Precise Measurement Fitting Algorithm Based on Priori Error Decomposition Weighting

Usually in industrial measurements, the size and posture of a workpiece instead of its absolute position are the most frequent factors of interest, but observation data include errors in both. To extract the errors that influence the size and posture of the workpiece, this paper presents an industrial fitting algorithm based on decomposition of priori errors to determine the optimal weight for each measured points. This algorithm executes prior error decomposition in multi-directions for spatial lines, planes, circles and spheres according to the shape of the workpiece and measuring postures. Using the decomposed value from the measurement error, the weight of each point was revaluated to do the fitting. Experimental results show that when compared with the equal-weighted fitting algorithm, this new algorithm improved fitting precision about 10%-17% for size and posture parameters.