Multiaxis motion control over a large travel range is still difficult for magnetic levitation positioning systems because the accuracy of the magnetic force model deteriorates when both translation and rotation of the moving part take place. To overcome this issue, this article studies a dynamic force- and torque-regulating method based on the numerical wrench model and validates it with a 6-degree-of-freedom magnetically levitated (maglev) rotary table. In each sampling cycle, the current-wrench transformation matrix is determined by the numerical wrench model, and then it is employed to decide the exciting currents for each coil phase by solving the overdetermined equation. A field programmable gate array is used as the processor to meet the real-time requirement of the control system by taking advantage of the fine-grained parallelism in the numerical wrench model. A high-level synthesis tool, general intellectual property core, a very-high-speed integrated circuit hardware description language component, and a MicroBlaze processor are employed simultaneously to develop the controller. By utilizing the numerical regulator, the maglev rotary table realizes stable multiaxis motions. Comparative experiments with the existing analytical method illustrate that the numerical wrench model improves the static decoupling effect, decreases the positioning fluctuations, and enlarges the travel range.

The magnetically levitated rotary table (MLRT) has the potential to be applied in many industrial applications due to its excellent rotation ability and good positioning accuracy. This paper presents the prototype of a novel MLRT with a simple and compact structure. The table employs a circular permanent magnet (PM) array and a circular coil array to generate large scale rotation around the vertical axis and limited movement in the other directions. This MLRT has a higher current carrying capacity and simpler commutation law than the existing MLRT. Furthermore, different from the previous works, the magnetic force model presented in this paper is obtained by a novel analytical-numerical method that considers the clearances in the circular PM array. Thus, the predicted magnetic force and torque are more accurate than those from the existing modeling method. The translation and rotation of the table are measured by six laser sensors, which are distributed around and below the mover respectively. The PID controller is used to stabilize the system. Experiments related to the motion decoupling, trajectory tracking and load change are carried out to verify the motion performance of the novel rotary table. Compared with the existing magnetic force model, the experimental results illustrate that the novel magnetic force model, which considers the clearances in a circular PM array, is more suitable for motion control of the MLRT.

The existing simulation method for the control of linear or planar magnetically levitated actuators always ignores the characteristics of the real physical object, which deteriorates the accuracy of the simulation. In this work, the proposed emulator for the magnetically levitated actuator is developed to consider both the force characteristics and the control algorithm. To model the real controlled object, the mathematical model for 1D (one-dimensional) and 2D (two-dimensional) magnetic arrays is derived where the yaw angle is taken into consideration using the coordinate transformation. The solution of the mathematical model is compared with the commercial BEM (boundary element method) software and the measurements from a force and torque testing setup to highlight the accuracy of the proposed mathematical model. Compared with the traditional simulation method of the motion control systems founded on the simplified system transfer function, the proposed simulation method has higher consistency and is closer to reality. The accuracy and efficiency of the proposed magnetic force model are further verified by the emulator based on the numerical force model and the testing data of the real setup.