| To achieve the high precision required in contouring machining applications, advanced servo-control algorithms are continuously being sought. In the conventional approach, the servo-controller is designed for each moving axis independently, regardless of the motion of other axes. This uncoupled control approach may result in degradation of contouring performance, although good tracking performance for each individual axis is achieved. The principal objective of the thesis is to develop a control architecture called the "cross-coupling controller (CCC)" to improve contouring accuracy in multi-axis machining systems.; The CCC method is based on constructing a contour-error model in real time and utilizing it in the determination of the control law that reduces or eliminates the contour error. Many researchers have implemented the CCC method and proven its superiority for linear cuts on two-axis systems. The existing CCCs, however, are not effective in dealing with nonlinear contours. Moreover, cross-coupling controllers have never been developed for more than two axes. The current research extends the CCC method to general nonlinear contours and general multi-axis systems.; Two-axis, three-axis, and five-axis cross-coupling controllers for general nonlinear contours have been proposed in the thesis. The proposed 2-axis CCC includes an improved contour-error model and a PID control law. The 3-axis CCC extends the 2D contour-error model to a 3D model. For the 5-axis CCC, the elimination of the orientation error as well as the elimination of the contour error are chosen as the two controller objectives. Consequently, the 5-axis CCC consists of a contour-error model, an orientation-error model, and a PID control law.; In addition to the proposed CCC method, other servo-control algorithms, such as PID-controller, ZPETC, and IKF methods, have been evaluated based on their abilities in error-source rejection and their limitations in machine tool control. The evaluation, which is supported by simulations and experiments, shows that the CCC method provides the best contouring accuracy.; The proposed CCC architecture extends the design of interpolators in the following way: more complicated nonlinear contours can be accurately controlled by the CCC method, and therefore, we may design new interpolator algorithms to generate these contours, which are usually approximated by lines and/or circles on current CNC systems. Accordingly, multi-axis interpolators have been proposed, which are capable of interpolating general nonlinear contours. |
