Real-time tool path modification for machine tool contour error reduction

As the demand for high-speed machining of complex features using computerized numerical control machine tools has increased, various contour control algorithms have been proposed to minimize contour errors. Tool path modification is a practical method for reducing contour errors because it does not require the modification of control structure, which is critical for implementing algorithms in a commercialized controller. Although the accuracy of contour estimation is crucial to the performance of this method, contour estimation errors are unavoidable due to the inaccuracy of the feed drive model as well as unexpected disturbances during machine operation. Thus, in this study, a robust real-time tool path modification method using a state estimator is proposed for more accurate estimation. A state estimator based on a Kalman filter is constructed using a feed drive model that includes the inertia of the feed drive and friction forces because the Kalman filter is robust to errors caused by model uncertainty, which was not considered in off-line tool path modification method. In addition, an advanced real-time contour error estimation method that utilizes only interpolated data is proposed. The effectiveness of the proposed tool path modification method is verified by experiments comparing it to the conventional off-line tool path modification method.