Abstract: Precision machining and inspection of complex surfaces have been the focuses of digital manufacturing. In order to improve machining and inspection accuracy of complex surface components, an integrated methodology of on-line inspection and machining error decomposition-compensation on NC machine is proposed. The principle of the system integrated on-line inspection and error compensation is then presented. The machining errors of measured points are analyzed after the machined surface is machined. The method of machining error decomposition is put forward using the spatial statistical analysis. A deterministic surface is established based on bicubic B-spline surfaces. Through the spatial statistical analysis of the residual errors of a regression model, the machining errors are decomposed into systematic errors and random errors. According to the decomposed systematic errors, the numeric control codes are modified and the machining errors compensation can be conducted. A part with a complex surface is taken as the example to verify the effectiveness of the proposed methodology. After the on-line inspection, the error decomposition, the code modification and the error compensation for the part, the machining accuracy of the test surface is greatly improved. The inspection accuracy of the system is also verified by comparison of the inspection data with a coordinate measuring machine.

Key words: Complex surface, Error compensation, Machining error, On-line inspection, Spatial statistical analysis, Experiment, Fuzzy reasoning , intelligent recognition , Irregular parking spaces, Multi - sensor Information Fusion