To provide a method of predicting deformation resistance in hot rolling, by which the deformation resistance which is higher in accuracy as compared with a conventional method is predicted without preparing a global approximate model like the conventional method.

The deformation resistance is predicted on the basis of a data base preparing step where respective past actual data is stored as a data base by taking factors which affect the flow stress in the hot rolling and the flow stress as an explanatory variable and an objective variable respectively, a request point data input step where the data of the explanatory variable corresponding to the deformation resistance which is going to be predicted is input as the request point data, a neighborhood data selecting step where distance calculation of the data stored in the data base and the request point data is performed and the data which is shorter in the calculated distance is selected as the neighborhood data, a local model preparing step where a local model which is locally fit to the neighborhood of the request point is prepared on the basis of the selected neighborhood data, the prepared local model and the request point data.

COPYRIGHT: (C)2010,JPO&INPIT