Modeling and prediction of workpiece nonlinear physical quantities based on extended Kalman filter

For the workpiece from the tool withdrawal time until the lifting of the fixture constraint is called the in-situ state. In this process, the workpiece's state quantity changes with time, and the state quantity can be regarded as a time series physical quantity. Therefore, the physical state and geometric physical quantities in the in-situ stage can be characterized by time sequence through the deformation theory. Kalman filter is a kind of optimal estimation method that includes two processes: prediction and correction. In the process of prediction estimation, only the optimal estimation of the last time needs to be stored, which can reduce the operation cost and improve the prediction efficiency. Since the state quantity of the workpiece in situ becomes nonlinear, this paper uses the extended Kalman filter for characterization modeling and nonlinear prediction, combines the typical heat-force coupling model with the extended Kalman filter for representation, and finally obtains the modeling characterization and result prediction of the filter state quantity.