The substantiation of using the artificial neural networks to compensate the kinematic and dy-namic errors of parallel kinematics machine tool

A system for active control of the spatial position of the parallel kinematics machine tool was developed. The method of compensation of kinematic and dynamic errors of the parallel kinematics machine tool using the artificial neural networks trained on the base of the proportional-differential controller was considered. The approbation of the developed proportional-differential controller, the output of which is applied to train the artificial neural network based on the experimental data of the transition process that occurs when the trajectory of the platform is changed. The feasibility of using the artificial neural networks in the feedback system of parallel kinematics machine tools was argumented based on the analysis of variance of the results after its study.