Abstract: To improve circular interpolation accuracy of 2x/y gantry linear feed axes, an error compensation methodology is focused on in the work. The technique consists of three parts: circular error measurement, evaluation and compensation. The causes of the error generation during the interpolation process and its compensation difficulties are discussed. Laser interferometer with dynamic precision measurement function is introduced to obtain the dual feed axis coordinate data during the process. The roundness error is evaluated through calculation of the data with least square roundness evaluation method. A model based on least square support vector regression (LS-SVR) approach is established to learn and recognize the rule of the error. With the support vectors derived from the learning process, the magnitude for error compensation is acquired through the real-time regression calculation. In addition, the compensation output control strategy is constructed and a corresponding system is developed to realize the online control process. To illustrate the procedure of the combination approach, an experiment is conducted on the self-construction linear axis feeding platform with the developed system. The data from the experiment show that the technique can compensate the interpolation circular error. As a result, the study finds 68.7 percent roundness accuracy improvement with the online compensation method.

Key words: Error compensation, Feed drive system, Linear motor, Roundness error