Elucidation of processing mechanism by statistical analysis and optimization of multi-characteristics Doctoral course in Doctor of Business Administration, Mejiro University 3rd year Akira Ogawa Visiting professor at Keio University Masanori Takahashi
Data representing the processed shape is called profile data.Profile data is generally expressed in two-dimensional or three-dimensional orthogonal coordinates or polar coordinates.If the object to be evaluated is three-dimensional, multiple profile data on the surface and cross section can be obtained from one object.By analyzing profile data obtained from processing experiments carried out under an appropriate plan by linear or nonlinear regression, it is possible to clarify the causal relation between the processed shape and the processing factor.Furthermore, it is possible to elucidate the processing mechanism by combining it with unique technical information.If the mechanism is elucidated, consider multiple profile data as objective functions, optimize them by mathematical programming under appropriate constraints, and feed back the optimum conditions to the machining process, thereby obtaining multiple characteristics of machining Can be optimized.
In this presentation, we analyze machining shape profile data based on statistical model by nonlinear regression of JMP, find optimum solution by mathematical programming method, and propose methodology of optimization of machining process.
