Analysis of CMM dynamic measurement error based on kernel estimation

Abstract

To overcome the sources of dynamic measurement error which are intricate and influence each other so that makes it difficult to model by parameter identification method in coordinate measuring machine's dynamic measurement error model. This article use kernel estimation to analyze the complicated relationship between dynamic measurement error values and three dimensional coordinates and direct computer control parameters including positioning velocity, approaching distance and contacting velocity. In the model, the differences of the variable corresponding to the measurement errors are considered, and the dynamic measurement errors are predicted. The research results show that the predict effect of the kernel estimation regression model has been significantly improved. It is better than ordinary least squares regression and partial least squares regression model by considering the interaction.