Magnetic suspension workbench structure and control method thereof

Abstract

The invention discloses a six-degree-of-freedom magnetic suspension workbench structure and a control method thereof, belongs to the technical field of magnetic suspension workbenches, and aims to solve the problems that the stroke of a traditional workbench is difficult to expand, the modeling precision is low, and the influence of a coil edge effect is large. The magnetic suspension workbench structurally comprises a stator capable of being spliced with a PCB coil array, a rotor of a symmetric two-dimensional Halbach permanent magnet array, a data processing module with an information fusion function and an H-bridge-containing driving and real-time control system. According to the control method, a high-precision electromagnetic force model is established, a magnetic field area is divided, a compensation coefficient is introduced to correct an end effect, a pseudo-inverse method based on least square is adopted to decouple and calculate the size of a driving current, and the saturation suppression of the current and the dynamic gating of the coil are realized in combination with a current calculation result; the method can be applied to the fields of precision machining and micro-nano manufacturing, and high-precision long-stroke motion control is achieved.

Description

Technical Field

[0001] This invention belongs to the field of magnetic levitation workbench technology, specifically relating to a six-degree-of-freedom magnetic levitation workbench structure and its control method. Background Technology

[0002] In the field of precision machining and manufacturing, magnetic levitation drive has become a core technological pillar of high-end equipment due to its non-contact force transmission characteristics and high-precision positioning capabilities. However, traditional magnetic levitation worktables face many problems in practical applications, which severely restrict their performance in high-precision, long-stroke scenarios.

[0003] Traditional magnetic levitation stages typically employ a design where the stator is completely fixed to the platform. This design results in a fixed and limited mover stroke. In complex and precision machining scenarios, such as the large-area scanning exposure of wafers in semiconductor lithography machines and the high-precision milling of large aerospace components, the stage requires a flexible and adjustable long stroke, which traditional structures clearly cannot meet. Taking magnetic levitation stages using copper-wound coils as an example, the coil manufacturing process is complex, requiring manual or specialized equipment for winding. This not only leads to low production efficiency but also results in poor consistency in parameters such as coil size, number of turns, and winding density. This further limits the flexible adjustment of the stroke and significantly increases the manufacturing cost and production time of the coil, hindering large-scale applications.

[0004] Modeling accuracy is another major technical challenge for traditional magnetic levitation worktables. Because traditional worktables often use copper-wound coils, problems such as coil size deviations, turn count errors, and uneven winding inevitably arise during manufacturing. These factors make electromagnetic force modeling extremely difficult. Existing models struggle to accurately describe the electromagnetic force relationship between the mover and stator. In actual control, there is a significant deviation between the model's predicted values ​​and the actual electromagnetic force values, severely impacting the worktable's control accuracy and dynamic performance. This makes it unsuitable for applications requiring extremely high positioning accuracy, such as micro / nano lithography and precision optical component fabrication.

[0005] Edge-end effects are another significant technical hurdle that traditional magnetic levitation stages struggle to overcome. When the mover approaches the edge of the stator, the magnetic field distribution generated by the stator coils undergoes significant distortion; this distortion is known as the edge-end effect. The edge-end effect leads to deviations in electromagnetic force calculations, resulting in a substantial decrease in the stage's positioning accuracy in the edge region. It also disrupts the stage's motion stability, causing unnecessary vibrations and impacts. During long-stroke movements, the mover frequently passes through the stator edge region, and the effects of the edge-end effect persist throughout the entire process, further limiting the performance improvement of traditional magnetic levitation stages in high-precision, long-stroke applications.

[0006] To address the technical shortcomings of traditional magnetic levitation worktables, this invention proposes a six-degree-of-freedom magnetic levitation worktable structure and its control method, which features flexible stroke expansion, high modeling accuracy, effective suppression of edge effects, and superior control performance. The aim is to break through existing technical bottlenecks and meet the demand for high-precision long-stroke motion control in the high-end manufacturing field. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a six-degree-of-freedom magnetic levitation stage structure and its control method. The stage has the characteristics of flexible stroke expansion, high modeling accuracy, good edge effect suppression, and superior control performance, which can effectively meet the needs of high-precision long-stroke motion control in fields such as precision machining, microscopic imaging technology, micro-nano manufacturing, biomedical engineering, and industrial control.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] This invention provides a six-degree-of-freedom magnetic levitation stage structure, including a stator and a mover;

[0010] The stator includes a PCB coil array (5), which adopts a multi-layer stacked wiring structure (16) and is equipped with standardized mechanical and electrical interfaces. The PCB coil array (5) is provided with a solder mask layer (12), coil copper pour (13), stacked coil layer (16) and bottom driving circuit layer (17) from top to bottom. The PCB coil array (5) is also provided with an annular hole (14) and a mounting circular hole (15). The annular hole (14) is used to reduce the weight of the module, and the mounting circular hole (15) is used to fix the PCB coil array (5) on the optical panel (3) of the worktable structure.

[0011] The mover includes a symmetrical two-dimensional Halbach permanent magnet array (8), which includes a main magnet block (11) and a secondary magnet block (10) with alternating magnetization directions. In the two-dimensional Halbach permanent magnet array (8), the magnetization direction (9) of the permanent magnet blocks changes periodically in the two-dimensional plane, where "×" indicates that the magnetization direction is from outside the paper to inside the paper, and "·" indicates that the magnetization direction is from inside the paper to outside the paper. This arrangement concentrates the magnetic field on the side of the mover facing the stator, reducing the magnetic field leakage on the other side of the mover.

[0012] A preferred embodiment of the present invention further includes a sensor system, a data processing module, and a control system;

[0013] The sensor system includes at least 6 laser displacement sensors (4), wherein 3 laser displacement sensors (4) are arranged in the horizontal direction to obtain the x-axis displacement, y-axis displacement and z-axis rotation angle γ of the mover by triangulation; and 3 laser displacement sensors (4) are arranged in the vertical direction to measure the z-axis displacement, x-axis rotation angle α and y-axis rotation angle β of the search mover.

[0014] The data processing module includes a multi-sensor pose information fusion and calibration unit;

[0015] The control system includes modular hardware and a software system developed based on a real-time control platform. The modular hardware includes a drive module, a data acquisition module, and a power management module. The control system is used to receive the mover pose information output by the data processing module, calculate and output control signals to the PCB coil array (5) based on the electromagnetic force model (25).

[0016] In a preferred embodiment of the present invention, the sensor system further includes a capacitive sensor; the multi-sensor pose information fusion unit of the data processing module adopts a Kalman filter algorithm or a particle filter algorithm; the calibration unit calibrates the sensor output signal through a standard measuring device to eliminate the sensor's zero-point drift, sensitivity error, and installation error; the standard measuring device includes a standard gauge block and a laser interferometer.

[0017] In a preferred embodiment of the present invention, the driving module of the control system is an H-bridge driving module (23); the software system includes a data acquisition program, a high-precision magnetic model (25) calculation program and a multi-degree-of-freedom PID control program. The data acquisition program is used to set the sensor sampling frequency of 1kHz-10kHz and the range. The high-precision magnetic model (25) calculation program is used to output the electromagnetic force vector according to the motioner's pose.

[0018] This invention provides a control method for a six-degree-of-freedom magnetic levitation worktable structure. Using the six-degree-of-freedom magnetic levitation worktable structure described above, the control method includes the following steps:

[0019] Step 1: Establish a high-precision electromagnetic force model (25): By calculating the magnetization intensity of the two-dimensional Halbach permanent magnet array (8), and combining Maxwell's equations and Poisson's equations, the distribution law of magnetic flux density in space is derived. The nonlinear effect of the magnetic field and the boundary condition correction model are introduced to describe the electromagnetic force interaction relationship between the mover and the stator.

[0020] Step 2: Edge effect correction: The magnetic field action area under the two-dimensional Halbach permanent magnet array (8) is divided into a central region (18), a side region (19) and a corner region (20). For the central region (18), the electromagnetic force calculation formula based on the harmonic model is directly adopted. For the side region (19) and the corner region (20), experimentally calibrated edge effect compensation coefficients k1, k2, and k3 are introduced. The harmonic model calculation results are corrected in real time by combining the lookup table method. The ideal magnetic field data at different positions are pre-stored by the lookup table method.

[0021] Step 3: Decoupling control: Solve for the pseudo-inverse matrix G of the coil current-power transformation matrix G using the least squares method. + Based on the electromagnetic force vector F obtained in step 1, the formula I = G is used. + ×F calculates the excitation current vector I of the stator PCB coil array (5) to achieve six-degree-of-freedom decoupling control;

[0022] Step 4: Dynamic selection of coils: Based on the relative position of the mover and the PCB coil array (5), determine the projection area of ​​the mover on the PCB coil array (5), and only supply power to the PCB coil modules that can generate effective magnetic field coupling in and near the projection area, and de-energize the other PCB coil modules.

[0023] In a preferred embodiment of the present invention, in step 2, k1 is the compensation coefficient of the side region (19), and k2 and k3 are the compensation coefficients of the corner region (20) in different directions. The compensation coefficients are obtained by moving the mover to different positions of the side region (19) and the corner region (20), collecting the actual magnetic field strength with a high-precision magnetic field measuring device, comparing it with the theoretical value of the harmonic model, and using the least squares method to fit and obtain the optimal values ​​of k1, k2, and k3.

[0024] A preferred embodiment of the present invention further includes a current saturation suppression step: introducing a diagonal weight matrix Ω to monitor the current value of the PCB coil and the power amplifier status in real time; when the solved current of a certain coil exceeds the power amplifier saturation threshold, or the power amplifier reports saturation, the weight coefficient corresponding to that coil is exponentially multiplied according to the formula I = Ω × G. + ×(G×Ω×G + ) - The excitation current vector I is recalculated using 1×F, and the current is limited to the saturation threshold range.

[0025] In a preferred embodiment of the present invention, in step 4, the specific judgment logic of the coil dynamic selection is as follows: if the magnetic field coupling strength between the PCB coil and the two-dimensional Halbach permanent magnet array (8) is greater than the set threshold, and the magnetic field coupling curve is determined based on the pre-simulated magnetic field coupling curve, then the coil is determined to be a "coil that can generate effective force", and the control system outputs a selection signal to the driving circuit of the coil.

[0026] In a preferred embodiment of the present invention, when establishing a high-precision electromagnetic force model (25) in step 1, the fused pose information output by the data processing module is also combined with the data from the laser displacement sensor (4) and the capacitance sensor through Kalman filtering to obtain the model parameters for real-time fine-tuning, thereby further improving the accuracy of electromagnetic force calculation.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] (1) Flexible stroke expansion to adapt to multiple scenario requirements: The stator adopts a splicable PCB coil array (5). By adding or removing modules through standardized interfaces, the stroke can be adjusted as needed. It can meet the large stroke processing requirements of semiconductor lithography machines and adapt to small stroke scenarios such as microscopic imaging. There is no need to redesign the overall structure, which reduces the equipment adaptation cost.

[0029] (2) Optimization of magnetic field performance to improve levitation and driving capabilities: The mover is a symmetrical two-dimensional Halbach permanent magnet array. The magnetic field is concentrated on the working side and evenly distributed, reducing magnetic field leakage. Compared with traditional copper winding coils, the magnetic levitation force and driving torque are greater, the load capacity and acceleration performance are improved, and the magnetic field harmonic distortion is reduced, laying the foundation for high-precision control.

[0030] (3) High modeling accuracy and effective suppression of edge effects: A high-precision electromagnetic force model is established based on Maxwell's equations. Combined with the division of magnetic field regions (central region, side region, corner region) and experimental calibration compensation coefficients k1, k2, k3, and real-time correction by table lookup method, the accuracy of electromagnetic force calculation is greatly improved. The positioning deviation caused by edge effects is suppressed throughout the entire stroke, ensuring motion stability.

[0031] (4) Superior control performance and safe and efficient operation: The Moore-Penrose pseudo-inverse method is used to achieve six degrees of freedom decoupling and eliminate coupling interference; the weighted pseudo-inverse strategy is used to suppress current saturation and avoid power amplifier failure; the coil dynamic selection only supplies power to the effective coil, which reduces energy consumption and control calculation, improves dynamic response speed, and adapts to high-speed and high-precision processing requirements.

[0032] (5) Accurate and reliable measurement, strong system adaptability: The multi-sensor system (6 laser displacement sensors and 4+ capacitive sensors) combined with Kalman filter data fusion has high position and pose measurement accuracy and can correct sensor errors; the modular design of the system (H-bridge drive module, LabVIEW software) facilitates maintenance and function expansion, and can be adapted to many fields such as precision machining, micro-nano manufacturing, and biomedical engineering, with a wide range of applications. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments or prior art, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This application provides a schematic diagram of a six-degree-of-freedom magnetic levitation worktable structure, illustrating the layout and connection relationships of the PCB coils, the two-dimensional Halbach permanent magnet array, and other key components, demonstrating the worktable's six-degree-of-freedom motion capability. The components are: 1-vertical sensor fixing disk, 2-aluminum gantry frame, 3-optical panel, 4-laser displacement sensor, 5-PCB coil array, 6-laser sensor positioning seat, 7-stator positioning screw, and 8-two-dimensional Halbach permanent magnet array.

[0035] Figure 2 This is a schematic diagram of the structure of a two-dimensional Halbach permanent magnet array provided in this application embodiment. It details the arrangement and magnetization direction of the permanent magnet blocks to achieve high magnetic field density and good symmetry, thereby improving the magnetic levitation effect and modeling accuracy. In the diagram, 9 represents the magnetization direction of the magnet block (where "×" indicates from outside the paper to inside the paper, and "·" indicates from inside the paper to outside the paper), 10 represents the secondary magnet block, and 11 represents the primary magnet block.

[0036] Figure 3 This is a schematic diagram of the PCB coil array (5), a component of the stator in the magnetic levitation worktable provided in this application embodiment, wherein 12-solder mask layer, 13-coil copper pour, 14-annular hole, 15-mounting circular hole, 16-stacked coil layer, and 17-bottom driving circuit layer.

[0037] Figure 4 This is a schematic diagram of the permanent magnet array, which is a component of the mover in the magnetic levitation workbench provided in this application embodiment. In order to create an accurate and high-precision magnetic force model, the permanent magnet array is divided into four corner regions 20 (Corner1, Corner2, Corner3, Corner4), four side regions 19 (Side1, Side2, Side3, Side4), and a central region 18 (Center); wherein, 18 is the central region, 19 is the side region, and 20 is the corner region.

[0038] Figure 5The control system block diagram of the magnetic levitation worktable provided in this application embodiment includes a hardware component consisting of a drive module, a data acquisition module, and a power management module, as well as a software component including a data acquisition program, a magnetic model calculation program, and a PID control program. It illustrates the collaborative working relationship between these components to achieve precise control of the worktable. Specifically, 21-desired input pose, 22-gravity compensation, 23-H-bridge drive module, 24-magnetic levitation system including stator and mover, 25-high-precision magnetic model, and 26-laser displacement sensor for pose detection. Detailed Implementation

[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. The terms "upper," "lower," "front," "rear," "left," and "right," etc., used when describing the installation position or direction of the structure or components in this embodiment are based on the orientation shown in the accompanying drawings. They are merely for convenience of description, used to distinguish the relative positions of various components or directions, and do not represent the orientation of the system or functional components in this embodiment during use.

[0040] like Figures 1-5 As shown, this embodiment of the invention provides a six-degree-of-freedom magnetic levitation stage structure, including a stator, a mover, a sensor system, a data processing module, and a control system. The functional parts work together to achieve high-precision six-degree-of-freedom motion control of the stage, which can be applied to fields requiring high-precision long-stroke motion control, such as precision machining, microscopic imaging technology, micro-nano manufacturing, biomedical engineering, and industrial control.

[0041] The stator includes a PCB coil array 5, which employs a multi-layer stacked wiring structure 16. This optimizes the coil layout to reduce eddy current losses while ensuring uniform magnetic field distribution after module assembly. The PCB coil array 5 is equipped with standardized mechanical and electrical interfaces for easy installation and disassembly, and can form a continuous and stable magnetic field region. The multi-layer wiring structure of the PCB coil array 5 helps optimize magnetic field distribution, reduce eddy current losses, and improve work efficiency. Its compact design also makes the entire workbench structure lighter and more compact, facilitating installation and deployment in limited spaces. In this embodiment, the PCB coil array 5 is fixed to the optical panel 3 by stator positioning screws 7. The optical panel 3 is equipped with multiple laser sensor positioning seats 6.

[0042] The mover comprises a symmetrical two-dimensional Halbach permanent magnet array 8, located above the PCB coil array 5. The two-dimensional Halbach permanent magnet array 8 includes primary magnet blocks 11 and secondary magnet blocks 10 with alternating magnetization directions. This symmetrical two-dimensional Halbach permanent magnet array consists of multiple permanent magnet blocks with alternating magnetization directions. This specific arrangement creates a high-density magnetic field on the side of the mover facing the stator, improving magnetic field utilization and reducing magnetic field harmonic distortion, thereby enhancing the stability and dynamic response performance of the magnetic levitation force. Compared to traditional permanent magnet arrays, this structure can generate greater force under the same size and current conditions, thus improving the load capacity and acceleration performance of the worktable and meeting the dynamic response requirements of high-speed, high-precision machining tasks. Simultaneously, its symmetrical design helps reduce magnetic field harmonic distortion, further improving magnetic field quality and providing favorable conditions for precise magnetic levitation control.

[0043] The sensor system includes at least six laser displacement sensors 4. Three laser displacement sensors 4 are arranged horizontally to obtain the x-axis displacement, y-axis displacement, and rotation angle γ around the z-axis of the mover using triangulation. Three laser displacement sensors 4 are arranged vertically to measure the z-axis displacement, rotation angle α around the x-axis, and rotation angle β around the y-axis of the mover. A vertical sensor fixing disk 3 is positioned above the optical panel 3 to fix the aforementioned sensors.

[0044] The data processing module includes a multi-sensor pose information fusion and calibration unit to improve pose information accuracy. The control system includes modular hardware and a software system developed based on the LabVIEW platform. The modular hardware includes a driver module, a data acquisition module, and a power management module. The control system receives the mover pose information output by the data processing module, calculates and outputs control signals to the PCB coil array 5 based on the electromagnetic force model 25, and the software system achieves six-degree-of-freedom decoupled control by acquiring the mover pose information in real time and combining it with the electromagnetic force model to generate control signals.

[0045] The electromagnetic force model is constructed based on Maxwell's equations and Poisson's equations. It combines the magnetization intensity distribution law of the two-dimensional Halbach array, stores the magnetic field data at different locations obtained by pre-calculation or simulation through a lookup table method, and uses the magnetic field symmetry characteristics of the central region of the worktable to compensate the magnetic field data of the edge region to correct the edge effect, thereby achieving high-precision calculation of electromagnetic force across the entire stroke range.

[0046] like Figure 3As shown, the PCB coil array 5 is arranged from top to bottom as follows: solder mask layer 12, coil copper pour 13, stacked coil layer 16, and bottom drive circuit layer 17. The PCB coil array 5 also has an annular hole 14 and a mounting circular hole 15. The annular hole 14 is used to reduce the weight of the module and reduce the overall weight of the workbench. The mounting circular hole 15 is used to fix the PCB coil array 5 to the optical panel 3 of the workbench structure to ensure that the module is installed firmly and reliably.

[0047] The PCB coil array 5 design in this embodiment fundamentally improves upon the limitation of traditional magnetic levitation stage stator coils inflexible expansion, enabling on-demand expansion of the stage travel. The PCB coil array 5 consists of multiple independent PCB coil modules for flexible travel expansion. Each PCB coil module employs a multi-layer stacked wiring structure 16. By optimizing the coil layout, eddy current losses are effectively reduced, while ensuring uniform magnetic field distribution after module assembly, providing a favorable magnetic field environment for stable levitation and precise movement of the mover. Each PCB coil module has standardized mechanical and electrical interfaces. The mechanical interface uses a precision positioning pin and fastening screw design, while the electrical interface uses standardized connectors, facilitating installation and disassembly between modules. In practical applications, the stator length and width can be flexibly adjusted by increasing or decreasing the number of PCB coil modules according to the required travel size, thereby achieving customized expansion of the stage travel. For example, in semiconductor wafer processing scenarios that require long strokes, the stator length can be extended to several meters by splicing more PCB coil modules to meet the large-scale scanning requirements of the wafer; while in precision microscopic imaging scenarios with short strokes, the number of PCB coil modules can be reduced, the stator volume can be reduced, and the space utilization of the equipment can be improved.

[0048] The solder mask layer 12 uses a high-temperature resistant and wear-resistant insulating material, which can effectively protect the coil copper layer 13 from external environmental corrosion and prevent short circuits. The coil copper layer 13 uses high-purity oxygen-free copper material, which is made through precise photolithography and etching processes. It has an extremely low resistance value, which can reduce current loss and improve power utilization. The stacked coil layer 16 is composed of multiple layers of coil copper layer 13, which are isolated from each other by an insulating layer. By rationally designing the number of turns and winding direction of each layer of coil, the magnetic field distribution can be further optimized and the magnetic field strength enhanced. The bottom driving circuit layer 17 integrates the driving chip, protection circuit and other components required for coil driving, which can realize precise control and protection of coil current and ensure safe and stable operation of the coil.

[0049] like Figure 2As shown, in the two-dimensional Halbach permanent magnet array 8, the magnetization direction 9 of the permanent magnet blocks changes periodically in the two-dimensional plane. "×" indicates that the magnetization direction is from outside the paper to inside the paper, and "·" indicates that the magnetization direction is from inside the paper to outside the paper. This arrangement concentrates the magnetic field on the side of the mover facing the stator, reducing magnetic field leakage on the other side of the mover.

[0050] In this embodiment, the mover adopts a symmetrical two-dimensional Halbach permanent magnet array 8 structure. Through a specific magnet arrangement, this structure can generate a highly dense and uniformly distributed magnetic field on the side of the mover facing the stator, laying a solid foundation for the high-precision suspension and drive of the worktable, and enabling the worktable to have superior stability and dynamic performance during operation.

[0051] The symmetrical two-dimensional Halbach permanent magnet array 8 consists of multiple permanent magnet blocks with alternating magnetization directions, including a main magnet block 11 and a secondary magnet block 10, the specific structure of which is as follows: Figure 2 As shown, both the main magnet block 11 and the secondary magnet block 10 are made of high-performance neodymium iron boron permanent magnet material, which has extremely high remanence and coercivity, and can generate a strong magnetic field. The magnetization direction 9 of the permanent magnet blocks changes periodically and alternately in a two-dimensional plane. Through this specific arrangement, the magnetic field can be concentrated on the side of the mover facing the stator, which greatly increases the magnetic field density, while significantly reducing the magnetic field strength on the other side of the mover, reducing magnetic field leakage, and improving magnetic field utilization. That is, the magnetic field density is concentrated on one side of the mover to enhance the magnetic levitation force and driving torque.

[0052] The sensor system also includes a capacitive sensor; the multi-sensor pose information fusion unit of the data processing module adopts a Kalman filter algorithm or a particle filter algorithm, and the calibration unit calibrates the sensor output signal through standard measuring equipment to eliminate sensor zero drift, sensitivity error and installation error. The standard measuring equipment includes standard gauge blocks and a laser interferometer.

[0053] The sensor system in this embodiment is used to monitor the position and pose information of the mover in real time, providing data support for the accurate feedback of the control system. Its performance directly affects the control accuracy and stability of the worktable. The sensor system adopts a multi-sensor fusion design, including a laser displacement sensor 4, a capacitive sensor, and other optional high-precision displacement or angle sensors. Through data fusion of multiple sensors, the measurement accuracy and reliability of position and pose information can be effectively improved.

[0054] The laser displacement sensor 4 has advantages such as high measurement accuracy, fast response speed, and strong anti-interference ability. It is mainly used to measure the displacement and angle information of the mover. The laser displacement sensor 4 is mounted on the aluminum gantry frame 2 via the laser displacement sensor positioning seat 6. The aluminum gantry frame 2 is fixed on the optical panel 3. The optical panel 3 has extremely high flatness and stability, which can ensure the installation accuracy of the laser displacement sensor 4. To achieve comprehensive monitoring of the six degrees of freedom pose of the mover, the sensor system adopts a specific layout: three laser displacement sensors 4 are arranged in the horizontal direction, which can accurately obtain the displacement x, displacement y, and rotation angle γ of the mover in the plane through triangulation; three laser displacement sensors 4 are arranged in the vertical direction to measure the vertical displacement z of the mover and the rotation angle α around the x-axis and the rotation angle β around the y-axis. Through the coordinated work and data calculation of the six laser displacement sensors 4, the spatial pose state of the mover can be completely characterized, providing comprehensive and accurate pose feedback information for the control system.

[0055] Capacitive sensors, characterized by their small measurement range, high resolution, and insensitivity to ambient temperature, can complement laser displacement sensor 4, further improving the accuracy of pose information measurement in scenarios requiring extremely high precision. Furthermore, depending on the specific application requirements, other types of high-precision sensors, such as grating rulers and encoders, can be selected to meet the measurement needs of different scenarios.

[0056] The control system's drive module is an H-bridge drive module 23; the software system includes a data acquisition program, a high-precision magnetic model 25 calculation program, and a multi-degree-of-freedom PID control program. The data acquisition program is used to set the sensor sampling frequency (1kHz-10kHz) and range, and the high-precision magnetic model 25 calculation program is used to output the electromagnetic force vector based on the mover's pose.

[0057] The control system in this embodiment is the core control unit of the magnetic levitation worktable. It receives the mover pose information output by the data processing module, calculates and outputs control signals to the stator's PCB coil module based on a preset electromagnetic force model, and achieves precise six-degree-of-freedom control of the worktable. The control system adopts a modular design concept, consisting of a hardware system and a software system. The parts work together to ensure the stable and precise operation of the worktable.

[0058] The hardware system adopts a highly integrated modular design, mainly including a driver module, a data acquisition module, and a power management module. The driver module uses an H-bridge driver module 23, which enables precise control of the PCB coil current, including adjustment of parameters such as current amplitude, frequency, and phase. The H-bridge driver module 23 integrates high-performance power semiconductor devices, such as MOSFETs and IGBTs, offering advantages such as fast switching speed and low conduction losses, meeting the high-precision and high-dynamic-response requirements of the PCB coil for drive current. The data acquisition module collects pose information output from the sensor system and status information such as current and voltage of the PCB coil. Employing a high-speed, high-precision A / D converter, it achieves rapid and accurate acquisition of analog signals, providing real-time data support for the control algorithm of the software system. The power management module provides a stable and reliable power supply for the entire hardware system, including functions such as input power filtering, voltage regulation, and overcurrent protection, ensuring that each module can operate normally under different working conditions.

[0059] The software system, developed on the LabVIEW platform, possesses powerful data acquisition, processing, and control capabilities. It primarily comprises a data acquisition program, a magnetic model calculation program, and a PID control program. The data acquisition program controls the data acquisition module to collect state data from the sensor system and PCB coils, and preprocesses the acquired data, such as filtering and noise reduction, to ensure data quality. The magnetic model calculation program, based on a high-precision electromagnetic force model, calculates the required electromagnetic force according to the actual pose information of the mover, thereby determining the target current value for each PCB coil. The PID control program employs a multi-degree-of-freedom PID controller, generating corresponding control signals based on the deviation between the target current value and the actual current value, and sending them to the drive module to achieve closed-loop control of the PCB coil current. Furthermore, the software system features a human-machine interface. Through a visual operating interface, users can monitor the operating status of the workbench in real time, such as the mover's pose, coil current, and voltage parameters, and adjust control parameters according to actual needs, such as the proportional coefficient, integral coefficient, and derivative coefficient of the PID controller, to optimize the workbench's control performance.

[0060] The six-degree-of-freedom magnetic levitation stage structure also includes a sensor calibration and data fusion module, which monitors the pose information of the mover in real time through laser displacement sensors, capacitive sensors, and other optional high-precision displacement or angle sensors, and uses data fusion algorithms to improve the accuracy of the pose information; at the same time, a fault diagnosis and protection mechanism is set up to detect and respond to abnormal operating states in real time, ensuring the safe and stable operation of the system.

[0061] The novel magnetic levitation worktable described above includes a mover and a stator positioned below it. The mover comprises a permanent magnet array, and the stator comprises a PCB coil array. The permanent magnet array is distributed above the PCB coil array and is arranged in a two-dimensional Halbach array. The PCB coil array comprises coils integrated in a stacked structure. Each PCB coil module is mechanically connected by precision positioning pins and fastening screws. During assembly, the connectable PCB coil arrays are first assembled according to the required stroke to form the stator. During the assembly process, the coil positions of each module must be precisely aligned, and the precision of the mechanical interface and the fastening force must ensure that a continuous and uniform magnetic field region is formed after assembly, preventing magnetic field leakage and uneven magnetic force distribution. Through precise alignment and positioning, the assembled PCB coils can form a continuous and uniform magnetic field region, providing a foundation for the stable levitation of the mover. At the same time, gaps or misalignments between coil modules should be avoided to prevent magnetic field leakage and uneven magnetic force distribution, which would affect the performance of the worktable.

[0062] The assembled stator is fixed to the optical panel 3 of the worktable base, and its stability is ensured by bolts and other fasteners. Then, a symmetrical two-dimensional Halbach permanent magnet array is installed on the mover. During installation, the relative position between the permanent magnet array and the PCB coil must be precisely adjusted to ensure the mover can float stably in the magnetic field. Simultaneously, the mover is balanced to reduce vibration and swaying during operation and improve the stability of the worktable. During the balancing process, the position of the counterweight on the mover is adjusted to align the center of mass of the mover with the center of the magnetic field.

[0063] In the assembly of the magnetic levitation worktable, sensor installation is a crucial step. Based on the designed installation locations, multiple high-precision sensors are mounted on the worktable. These sensors include, but are not limited to, laser displacement sensors and capacitive sensors. Their function is to monitor the motion of the mover in real time, providing data support for precise feedback from the control system. Three sensors are arranged horizontally to obtain the mover's displacements x and y in the plane and its rotation angle γ around the z-axis using triangulation. Three sensors are arranged vertically to measure the mover's vertical displacement z and its rotation angles α and β around the x and y axes. Through the collaborative calculation of these six sensors, the spatial pose of the mover can be fully characterized. During sensor installation, it is essential to ensure that the sensor's measurement direction aligns with the mover's motion direction to ensure accurate capture of the mover's motion state. Simultaneously, a precision mechanical adjustment mechanism is used to fine-tune the sensor's position and angle to achieve optimal measurement results.

[0064] After sensor installation is complete, the next step is calibration. Standard measuring equipment is used to calibrate the sensor's output signal, ensuring the accuracy and reliability of the measurement data. During calibration, the sensor's output needs to be compared with known standards, and sensor parameters adjusted to eliminate deviations. Simultaneously, data from multiple sensors is fused to improve the accuracy and completeness of pose information. The data fusion process involves comprehensive analysis of data from different sensors, using algorithm optimization to improve the quality of the mover pose information.

[0065] Hardware connection is a crucial step in assembling the new magnetic levitation worktable. The drive module, data acquisition module, and power management module need to be connected via standardized interfaces to ensure normal signal transmission between them. During connection, the quality of the cables should be carefully inspected to ensure they are free from damage and aging, and the connections must be secure to prevent signal interruptions or instability due to poor contact, which could affect subsequent control performance. Furthermore, attention must be paid to cable layout to avoid mutual interference, especially maintaining an appropriate distance between signal and power lines to reduce the impact of electromagnetic interference on signal transmission.

[0066] After completing the hardware connection, a comprehensive debugging of the entire hardware system is required. First, use tools such as a multimeter to check the continuity of the coils, determining if there are any short circuits or open circuits by measuring the coil resistance. If an abnormal coil resistance value is found, immediately check for inter-turn short circuits or other faults, and repair or replace any damaged coils promptly. Next, adjust the magnitude and direction of the drive current, observing the levitation state and movement of the mover. By repeatedly adjusting the drive parameters, such as the current amplitude, frequency, and phase, ensure the mover can stably levitate within the predetermined position range and move along the expected trajectory. During debugging, closely monitor the smoothness of the mover's movement and response speed to ensure the worktable maintains good dynamic performance and stability under various operating conditions.

[0067] In configuring the software system of the new magnetic levitation stage, the first step is to configure the data acquisition program on the LabVIEW platform. Based on the characteristics of the selected sensors, key parameters such as sampling frequency and range are set to achieve efficient acquisition and processing of sensor data. The sampling frequency setting needs to comprehensively consider the sensor's response speed and data processing capabilities to ensure real-time capture of the mover's pose changes. The range setting must be determined based on the sensor's maximum measurement range and the actual motion range of the stage to avoid signal overflow and measurement errors. Simultaneously, the established high-precision magnetic model and decoupled control algorithm are seamlessly integrated into the software system. By inputting the desired mover pose, the required current values ​​for each coil are accurately calculated, and control signals are promptly sent to the drive module to achieve precise control of the mover.

[0068] During the operation of the software system, real-time monitoring of the mover's position and the current feedback of each coil is crucial to ensuring control accuracy. Based on the feedback data, the control system is adjusted in a closed loop to optimize control parameters, thereby improving the control accuracy and dynamic performance of the worktable. This closed-loop control mechanism effectively handles various disturbances and changes that may occur during worktable operation, such as load variations and external vibrations, ensuring that the mover always moves stably along the expected trajectory. Furthermore, various protection functions, such as overcurrent protection and overvoltage protection, can be set in the software system to ensure the safety of the worktable during operation. These protection functions can react quickly upon detecting abnormalities, cutting off power or adjusting control signals to prevent hardware damage and accidents.

[0069] A series of tests were conducted using a newly constructed six-degree-of-freedom magnetic levitation workbench. First, a step response test was performed. A step signal was input to the workbench, and the response speed and overshoot of the mover in each degree of freedom direction were observed to evaluate the workbench's positioning resolution and dynamic response characteristics. Then, ramp trajectory tracking and circular trajectory tracking tests were conducted. Ramp signals and circular trajectory signals were applied to the translational and rotational directions of the mover, respectively. The actual motion trajectory of the mover was recorded, and the tracking error was calculated to verify the workbench's positioning and tracking capabilities and anti-interference ability under multi-degree-of-freedom cooperative control. Based on the test results, the high-precision magnetic model was further revised and improved, and the parameters of the control algorithm were optimized and adjusted, such as adjusting the parameters of the PID controller and optimizing the weight matrix in the weighted pseudo-inverse method, to improve the overall performance of the workbench.

[0070] To ensure the long-term stable operation of the magnetic levitation stage, regular maintenance is essential. Maintenance primarily involves cleaning the PCB coils to prevent the accumulation of dust and dirt from affecting their heat dissipation and insulation performance. When cleaning, use a specialized electronic equipment cleaner and a soft brush to gently remove dust from the coil surface, avoiding the use of excessive liquid to prevent damage. Simultaneously, it is necessary to periodically check for changes in the magnetism of the permanent magnet array, which can be detected using specialized equipment such as a gaussmeter. If a weakening of the magnetism is detected, the corresponding permanent magnet should be replaced promptly to maintain the magnetic field strength and ensure stable levitation of the mover. Furthermore, regular calibration of the sensors is also a crucial part of maintenance. During calibration, standard measuring equipment should be used to calibrate the sensor output signals to ensure the accuracy and reliability of the measurement data.

[0071] Besides regular maintenance, establishing a comprehensive fault diagnosis system is equally crucial for ensuring the stable operation of the magnetic levitation worktable. This system can monitor the worktable's operating status in real time, including key parameters such as the mover's posture, coil current, and voltage. Once an anomaly is detected, such as excessive posture deviation or abnormal current, the system will immediately trigger an alarm and automatically record relevant data. Technicians can then quickly troubleshoot and resolve the problem based on this data. By analyzing the fault symptoms and related data, the cause of the fault can be accurately determined, such as hardware failure, software failure, or electromagnetic interference. For different causes, corresponding solutions can be implemented, such as replacing damaged hardware components, fixing software vulnerabilities, or optimizing electromagnetic shielding measures. These measures ensure that the worktable can quickly return to normal operation, effectively reducing downtime.

[0072] The novel magnetic levitation stage of this invention is not only applicable to conventional precision machining and microscopic imaging, but can also be extended to more application scenarios through appropriate adjustments and optimizations. For example, in the field of micro-nano manufacturing, the stage's positioning accuracy and stability can be further improved to meet the extremely high requirements for machining precision at the micro-nano scale. In biomedical engineering, through compatibility design with biological samples and aseptic processing, it can be used for the precise manipulation and observation of biological samples. Furthermore, the stage's control system can be upgraded and its functions expanded according to different application scenarios, such as adding remote control functions and collaborative control functions with other devices, to improve the stage's versatility and flexibility and meet diverse needs.

[0073] This invention also provides a control method for a six-degree-of-freedom magnetic levitation worktable structure. Using the six-degree-of-freedom magnetic levitation worktable structure described above, the control method includes the following steps:

[0074] Step 1: Establish a high-precision electromagnetic force model 25: By calculating the magnetization intensity of the symmetrical two-dimensional Halbach permanent magnet array 8, and combining Maxwell's equations and Poisson's equations, the distribution law of magnetic flux density in space is derived. The nonlinear effect of the magnetic field and the boundary condition correction model are introduced to describe the electromagnetic force interaction relationship between the mover and the stator.

[0075] In step 1, when establishing the high-precision electromagnetic force model 25, the fused pose information output by the data processing module is also combined with the data from the laser displacement sensor 4 and the capacitance sensor through Kalman filtering to perform real-time fine-tuning of the model parameters, further improving the accuracy of electromagnetic force calculation.

[0076] In this embodiment, an electromagnetic force model was established to ensure precise control of the magnetic levitation worktable. A high-precision electromagnetic force model was studied and established in depth. This model can accurately describe the electromagnetic force interaction between the mover and the stator, providing reliable theoretical support for the formulation of control strategies for the control system.

[0077] The electromagnetic force model is established based on fundamental electromagnetic theories such as Maxwell's equations and Poisson's equations. Combining this with the magnetization distribution of a symmetrical two-dimensional Halbach permanent magnet array 8, the spatial distribution of magnetic flux density is obtained through rigorous theoretical derivation and numerical calculations. First, the magnetization of the two-dimensional Halbach permanent magnet array 8 is calculated in detail. Based on the magnetization direction 9 and dimensional parameters of the permanent magnet blocks, a mathematical model of the magnetization intensity is established. Then, based on Maxwell's equations, the governing equations of the magnetic vector potential are derived. Combining the boundary conditions of the two-dimensional Halbach permanent magnet array 8, numerical calculation methods such as finite element analysis and analytical methods are used to solve for the spatial distribution of the magnetic vector potential. Finally, based on the relationship between magnetic flux density and magnetic vector potential, the spatial distribution of magnetic flux density is calculated. Furthermore, based on the Lorentz force formula, the formula for calculating the electromagnetic force between the mover and stator is derived.

[0078] In establishing the electromagnetic force model, various factors that may occur in actual working scenarios were fully considered, such as the nonlinear effects of the magnetic field and the influence of boundary conditions, and the model was continuously improved and corrected. The nonlinear effects of the magnetic field mainly originate from the hysteresis and saturation characteristics of the permanent magnet material. By introducing hysteresis and saturation models, the nonlinear effects of the magnetic field are described and corrected, improving the accuracy of the model. The influence of boundary conditions mainly includes factors such as the geometric boundaries of the stator and mover, and air gaps. By accurately setting the boundary conditions, it is ensured that the model can accurately reflect the actual magnetic field distribution.

[0079] To further improve the computational accuracy of the electromagnetic force model in the edge region, this invention introduces a lookup table method as an auxiliary means. Ideal magnetic field data under different positions and operating conditions are obtained through pre-calculation or simulation and stored in a database. During actual operation, the six-degree-of-freedom pose information of the mover is obtained through the sensor system. After the data processing module processes this pose information, the control system quickly retrieves the corresponding ideal magnetic field data from the database based on the actual pose of the mover and compares it with the actually measured magnetic field data to correct the computational results of the electromagnetic force model in real time. This correction method not only effectively improves the computational accuracy of the electromagnetic force model in the edge region but also avoids complex mathematical calculations, ensuring the real-time performance and efficiency of the control system.

[0080] Step 2: Edge effect correction: The magnetic field action area below the symmetrical two-dimensional Halbach permanent magnet array 8 is divided into a central region 18, a side region 19, and a corner region 20. For the central region 18, the electromagnetic force calculation formula based on the harmonic model is directly applied. For the side region 19 and the corner region 20, experimentally calibrated edge effect compensation coefficients k1, k2, and k3 are introduced. The harmonic model calculation results are corrected in real time using a lookup table method. The lookup table method pre-stores ideal magnetic field data at different locations.

[0081] In step 2, the edge effect compensation coefficient k1 corresponds to the side region 19, and k1 is the compensation coefficient of the side region 19. k2 and k3 are the compensation coefficients of the corner region 20 in different directions. The compensation coefficients are obtained by moving the mover to different positions in the side region 19 and the corner region 20, collecting the actual magnetic field strength with a high-precision magnetic field measurement device, comparing it with the theoretical value of the harmonic model, and using the least squares method to fit and obtain the optimal values ​​of k1, k2, and k3.

[0082] To address the unavoidable edge effect problem in the long-stroke motion of a magnetic levitation worktable, this invention proposes an efficient correction strategy. By utilizing the symmetry of the magnetic field distribution of the two-dimensional Halbach permanent magnet array, the magnetic field distortion caused by the edge effect is compensated and corrected, ensuring that the worktable maintains a stable levitation effect and precise motion control throughout its entire stroke range.

[0083] First, based on the magnetic field distribution characteristics of the two-dimensional Halbach permanent magnet array 8, the magnetic field action area below it is divided into different regions, specifically including a central region 18, four side regions 19, and four corner regions 20, as follows: Figure 4 As shown. The central region 18 is located at the center of the magnetic field area. The magnetic field distribution in this region is uniform and stable, and it is minimally affected by the edge effect. The lateral regions 19 are located around the central region 18, corresponding to the four lateral directions of the mover. The magnetic field distribution in this region is significantly affected by the edge effect, and the magnetic field will exhibit a certain degree of attenuation and distortion. The corner regions 20 are located at the four corners of the magnetic field area. The magnetic field distribution in this region is most severely affected by the edge effect, and the magnetic field attenuation and distortion are the greatest.

[0084] Different electromagnetic force calculation methods are used for different regions based on their magnetic field characteristics:

[0085] For the central region 18, since its magnetic field distribution is uniform and stable, it is minimally affected by edge effects. Therefore, the electromagnetic force calculation formula based on the harmonic model is directly used for calculation. The harmonic model can accurately describe the electromagnetic force relationship in a uniform magnetic field, has high calculation accuracy, and the calculation process is simple, which can meet the real-time requirements of the control system.

[0086] For the side region 19 and the corner region 20, their magnetic field distribution is significantly affected by the edge effect, and directly using the harmonic model for calculation will produce large errors. Therefore, this invention introduces experimentally calibrated edge effect compensation coefficients k1, k2, and k3 to correct the calculation results of the harmonic model. Here, k1 is the compensation coefficient for the side region 19, and k2 and k3 are the compensation coefficients for the corner region 20 in different directions, respectively. These coefficients characterize the magnetic field attenuation and nonlinear characteristics caused by the edge effect.

[0087] The edge effect compensation coefficients k1, k2, and k3 were obtained through extensive experimental calibration. During the experiments, the mover was moved to different positions in the side region 19 and the corner region 20, and the actual magnetic field strength at each position was measured using high-precision magnetic field measurement equipment. This measured magnetic field strength was then compared with the theoretical magnetic field strength calculated by the harmonic model. Based on the comparison results, data fitting methods such as the least squares method were used to determine the optimal values ​​of the edge effect compensation coefficients k1, k2, and k3. The compensation coefficients obtained through experimental calibration accurately reflect the edge effect characteristics of different regions, ensuring the effective correction of the harmonic model calculation results.

[0088] In practical applications, the control system determines the current magnetic field region of the mover based on its real-time pose information. If the mover is located in the central region 18, the electromagnetic force is directly calculated using the harmonic model. If the mover is located in the side region 19 or the corner region 20, the electromagnetic force calculation result is corrected by multiplying the harmonic model calculation result by the corresponding edge effect compensation coefficients k1, k2, k3. This regional correction method effectively suppresses the influence of edge effects on the accuracy of electromagnetic force calculation, ensuring that the worktable obtains accurate electromagnetic force control throughout its entire stroke range, thereby improving the positioning accuracy and motion stability of the worktable.

[0089] This invention introduces a lookup table method as an auxiliary means. By pre-storing ideal magnetic field data under different positions and operating conditions, and processing and comparing the six-degree-of-freedom pose information obtained by sensors during actual operation, the stored data is quickly retrieved and real-time corrections are made, thereby significantly improving the calculation accuracy of the high-precision magnetic model in the edge region. This correction strategy not only effectively solves the impact of edge effects on the control accuracy of the worktable, but also avoids complex mathematical calculations, ensuring the real-time performance and efficiency of the control system, enabling the worktable to maintain a stable suspension effect and precise motion control throughout its entire stroke range.

[0090] Step 3: Decoupling control: The Moore-Penrose pseudo-inverse method is used to solve for the pseudo-inverse matrix G of the coil current-power transformation matrix G. + Based on the electromagnetic force vector F obtained in step 1, the formula I = G is used. +×F calculates the excitation current vector I of the stator PCB coil array 5 to achieve six-degree-of-freedom decoupling control.

[0091] The decoupling control algorithm is as follows:

[0092] Because of the strong coupling relationship among the six degrees of freedom of the magnetic levitation stage, the motion control of one degree of freedom can interfere with the other degrees of freedom, seriously affecting the control accuracy and dynamic performance of the stage. To solve this problem, this invention uses the advanced Moore-Penrose pseudo-inverse method to solve for the coil current, thereby achieving decoupled control of the magnetic levitation stage.

[0093] The Moore-Penrose pseudo-inverse method is a commonly used matrix pseudo-inverse solution method. It can find an optimal approximate inverse matrix when the matrix is ​​not invertible, thus solving a system of linear equations. In the control of a magnetic levitation platform, a linear relationship model between coil current and electromagnetic force is first established. This model can be represented as a matrix equation: F = G × I, where F is the six-degree-of-freedom electromagnetic force vector, G is the current-to-force transformation matrix, and I is the excitation current vector of the PCB coil (I represents the excitation current vector of the n coils below the mover). Since the current-to-force transformation matrix G is usually non-square and non-invertible, the coil current vector I cannot be directly solved by matrix inversion. Therefore, the Moore-Penrose pseudo-inverse method is used to solve for the pseudo-inverse matrix G of the current-to-force transformation matrix G. + (G + (where G is the pseudo-inverse of the current-dynamic transformation matrix), and then through the formula I = G + ×F is used to solve for the coil current vector I.

[0094] By solving for the coil current using the Moore-Penrose pseudo-inverse method, independent control of the electromagnetic forces in all six degrees of freedom can be achieved, effectively suppressing coupling interference between the degrees of freedom. For example, when controlling the mover to translate along the x-axis, the decoupling control algorithm can accurately calculate the current of each PCB coil required to achieve this translation, avoiding interference with the mover's motion in other degrees of freedom such as the y-axis and z-axis, thereby improving the control accuracy and dynamic performance of the worktable.

[0095] The current saturation suppression algorithm is as follows:

[0096] During the movement of the mover, factors such as load changes and external interference may cause the solution current of some PCB coils to exceed the saturation threshold of the power amplifier, thus causing the power amplifier to enter a saturation state and fail to output the required current, severely affecting the control performance of the worktable and even potentially leading to worktable malfunction. To solve this problem, this invention proposes an improved weighted pseudo-inverse solution strategy, which avoids saturation by dynamically adjusting the current distribution of each coil.

[0097] The weighted pseudo-inverse solution strategy, based on the Moore-Penrose pseudo-inverse method, introduces a diagonal weight matrix Ω (Ω is a diagonal weight matrix). By adjusting the values ​​of each element in the weight matrix Ω, the weights of each coil current in the solution process are changed, thereby achieving dynamic allocation of the coil current. The mathematical model of the weighted pseudo-inverse solution is as follows:

[0098]

[0099] When the solution current value I of a certain coil is detected i (i = 1, 2, ..., n, where n is the total number of coils below the mover) exceeds the saturation threshold I of the power amplifier. sat Or, when the power amplifier feedback is saturated, the weighting coefficient Ω in the diagonal weighting matrix Ω corresponding to the coil is... ii The mathematical expression for exponential multiplication is:

[0100] in, This refers to the weighting coefficient of the coil before adjustment. The adjusted weighting coefficients are denoted by k, which is the exponential multiplication factor (set according to the system's dynamic response requirements and power amplifier characteristics, usually a constant between 0.5 and 2). The initial state of the weighting matrix Ω is the identity matrix.

[0101] When the current value of a certain coil exceeds the saturation threshold of the power amplifier or the power amplifier reports saturation, the weighting coefficient corresponding to that coil is increased exponentially.

[0102] Formula (1) is a linear relationship model between coil current and electromagnetic force; Formula (2) is a weighted pseudo-inverse solution formula for calculating coil current. By introducing the weight matrix Ω, the distribution ratio of each coil current can be adjusted; Formula (3) is a dynamic adjustment rule for the weight coefficient, which is used to suppress coil current saturation.

[0103] The weight matrix Ω is initialized to an identity matrix, and the weighted pseudo-inverse solution strategy is the same as the traditional Moore-Penrose pseudo-inverse method. During the workbench operation, the solved current values ​​of each PCB coil and the operating status of the power amplifier are monitored in real time. When the solved current value of a coil exceeds the saturation threshold of the power amplifier, or when the power amplifier reports saturation, the weight coefficient corresponding to that coil is exponentially increased. This exponential increase reduces the weight of that coil in the current solution process, decreasing the current allocation to that coil and thus limiting its current within the saturation threshold range. Simultaneously, since the weight matrix Ω is a diagonal matrix, adjusting the weight coefficient of one coil will not affect the current allocation of other coils, ensuring that other coils can operate normally.

[0104] By dynamically adjusting the weighting coefficients, the current of each coil can be strictly constrained within the specified saturation threshold range while ensuring the accuracy of the driving force. This effectively avoids the saturation phenomenon of the power amplifier and ensures that the workbench can operate stably under different working conditions.

[0105] Step 4: Dynamic Coil Selection: Based on the relative position of the mover and the PCB coil array 5, determine the projection area of ​​the mover on the PCB coil array 5, and only supply power to the PCB coil modules that can generate effective magnetic field coupling in and near the projection area, while de-energizing the remaining PCB coil modules.

[0106] In step 4, the specific judgment logic for the dynamic selection of the coil is as follows: if the magnetic field coupling strength between the PCB coil and the two-dimensional Halbach permanent magnet array 8 is greater than the set threshold, based on the pre-simulated magnetic field coupling curve, the coil is determined to be a "coil that can generate effective force", and the control system outputs a selection signal to the driving circuit of the coil.

[0107] The control method for the six-degree-of-freedom magnetic levitation stage structure also includes a current saturation suppression step: a diagonal weight matrix Ω is introduced, initially an identity matrix, and the current value of the PCB coils and the power amplifier status are monitored in real time; when the solution current of a certain coil exceeds the power amplifier saturation threshold, or the power amplifier reports saturation, the weight coefficient corresponding to that coil is exponentially multiplied according to the formula I = Ω × G. + ×(G×Ω×G + ) - The excitation current vector I is recalculated using 1×F, and the current is limited to the saturation threshold range.

[0108] The dynamic gating algorithm for coils is as follows:

[0109] Because the stator of the magnetic levitation stage uses a modular PCB coil array 5 with a large number of coils, continuously powering all coils would not only waste a lot of energy but also increase the computational burden on the control system, affecting the system's dynamic performance. To solve this problem, this invention proposes a dynamic coil selection strategy based on the relative positional relationship between the permanent magnet array and the PCB coils, selectively powering the coils that can generate effective force.

[0110] The core idea of ​​the coil dynamic gating algorithm is to determine which PCB coils have effective magnetic field coupling with the two-dimensional Halbach permanent magnet array 8 of the mover based on the real-time pose information of the mover, that is, which can generate electromagnetic force that contributes to the motion control of the mover. Then, only these coils are powered, while other coils that cannot generate effective force are de-energized.

[0111] The specific implementation process is as follows: First, a model of the relative positional relationship between the permanent magnet array and the PCB coils is established. This model can determine the projection area of ​​the mover on the stator PCB coil array 5 based on the mover's pose information. Then, based on the principle of magnetic field coupling, it is determined whether the PCB coils within the projection area can generate effective magnetic field coupling with the permanent magnet array. Typically, only the PCB coils located directly below and within a certain range of the mover can generate strong magnetic field coupling with the permanent magnet array, thereby generating effective electromagnetic force; while the PCB coils located outside this range have weak magnetic field coupling with the permanent magnet array, and the electromagnetic force they generate contributes very little to the mover's motion control, even to the point of being negligible.

[0112] Based on the above judgment results, the control system generates a coil selection signal, supplying power only to the PCB coils that can generate effective force, and de-energizing the other coils. For example, when the mover is located at the center of the stator PCB coil array 5, only the PCB coils directly below the mover are powered; when the mover moves along the x-axis, the selected PCB coils are dynamically adjusted according to the change in the mover's position, turning off the coils in the area the mover leaves and turning on the coils in the area the mover is about to enter, ensuring that only the coils that generate effective magnetic field coupling with the mover are always in working condition.

[0113] By employing a dynamic coil gating strategy, unnecessary coil power supply can be significantly reduced, improving the system's energy efficiency. Simultaneously, by decreasing the number of coils involved in control, the computational burden on the control system is reduced, increasing the processing speed of the control algorithm. This further enhances the dynamic performance of the worktable, enabling high-precision position tracking and attitude adjustment of the mover.

[0114] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A six degree of freedom magnetic levitation table structure, characterized by, The stator comprises a PCB coil array (5) which adopts a multi-layer stacked wiring structure (16) and is configured with a standardized mechanical interface and an electrical interface; the PCB coil array (5) is sequentially provided with a solder mask layer (12), a coil copper layer (13), a stacked coil layer (16) and a bottom driving circuit layer (17) from top to bottom; the PCB coil array (5) is further provided with a ring-shaped hole (14) and a mounting circular hole (15), the ring-shaped hole (14) is used for reducing the weight of the module, and the mounting circular hole (15) is used for fixing the PCB coil array (5) to an optical panel (3) of a workbench structure. The mover comprises a symmetric two-dimensional Halbach permanent magnet array (8) which comprises main magnet blocks (11) and secondary magnet blocks (10) with alternating magnetization directions; in the two-dimensional Halbach permanent magnet array (8), the magnetization directions (9) of the permanent magnet blocks periodically alternate in a two-dimensional plane, wherein "×" represents that the magnetization direction points from the paper surface to the paper surface, and "·" represents that the magnetization direction points from the paper surface to the paper surface; through this arrangement, the magnetic field is concentrated on the side of the mover facing the stator, and the magnetic field leakage on the other side of the mover is reduced. Further comprising a sensor system, a data processing module and a control system; 2. The six degree-of-freedom magnetic levitation table structure according to claim 1, characterized in that, The sensor system at least includes six laser displacement sensors (4), wherein three laser displacement sensors (4) are arranged in the horizontal direction to obtain the x-axis displacement, y-axis displacement and rotation angle γ around the z-axis of the mover by triangulation; three laser displacement sensors (4) are arranged in the vertical direction to measure the z-axis displacement, rotation angle α around the x-axis and rotation angle β around the y-axis of the search mover; The data processing module comprises a multi-sensor pose information fusion and calibration unit; The control system comprises a modular hardware and a software system developed based on a real-time control platform, the modular hardware comprises a driving module, a data acquisition module and a power management module; the control system is used for receiving the mover pose information output by the data processing module, calculating and outputting control signals to the PCB coil array (5) based on an electromagnetic force model (25). The sensor system further includes a capacitive sensor; the multi-sensor pose information fusion unit of the data processing module adopts a Kalman filtering algorithm or a particle filtering algorithm, the calibration unit calibrates the sensor output signal by using a standard measuring device to eliminate the zero point drift, sensitivity error and installation error of the sensor, and the standard measuring device includes a standard gauge block and a laser interferometer.

3. The six degree-of-freedom magnetic levitation table structure according to claim 2, characterized in that, The driving module of the control system is an H-bridge driving module (23); the software system includes a data acquisition program, a high-precision magnetic force model (25) calculation program and a multi-degree-of-freedom PID control program, the data acquisition program is used for setting the sensor sampling frequency of 1kHz-10kHz and the range, and the high-precision magnetic force model (25) calculation program is used for outputting an electromagnetic force vector according to the mover pose.

4. The six degree-of-freedom magnetic levitation table structure according to claim 3, characterized in that, ​ 5. A method of controlling a six-degree-of-freedom magnetic levitation table structure, characterized by, The control method comprises the following steps: Step 1: Establishing a high-precision electromagnetic force model (25): by calculating the magnetization intensity of the two-dimensional Halbach permanent magnet array (8), combining Maxwell's equations and Poisson's equation to derive the distribution of magnetic flux density in space, introducing a magnetic field nonlinear effect and boundary condition correction model to describe the electromagnetic force interaction relationship between the mover and the stator; Step 2: End effect correction: dividing the magnetic field action area below the two-dimensional Halbach permanent magnet array (8) into a central region (18), a side region (19) and an angle region (20), directly using the electromagnetic force calculation formula based on the harmonic model for the central region (18), introducing the experimentally calibrated end effect compensation coefficients k1, k2 and k3 for the side region (19) and the angle region (20), and combining the look-up table method to real-time correct the harmonic model calculation results, and the look-up table method pre-stores ideal magnetic field data at different positions; Step 3: decoupling control: the pseudo-inverse matrix G of the coil current-power conversion matrix G is solved by least square method + , the excitation current vector I of the stator PCB coil array (5) is calculated by the formula I = G + × F according to the electromagnetic force vector F obtained in step 1, and six-degree-of-freedom decoupling control is realized; Step 4: Coil dynamic gating: according to the relative position of the mover and the PCB coil array (5), determining the projection area of the mover on the PCB coil array (5), and only supplying power to the PCB coil modules in and near the projection area which can produce effective magnetic field coupling, and turning off the rest of the PCB coil modules.

6. The method of controlling a six-degree-of-freedom magnetic levitation table structure according to claim 5, wherein, In step 2, k1 is the compensation coefficient of the side region (19), and k2 and k3 are the compensation coefficients of the angle region (20) in different directions; the compensation coefficients are obtained by: moving the mover to different positions in the side region (19) and the angle region (20), collecting the actual magnetic field intensity with a high-precision magnetic field measuring device, comparing with the harmonic model theoretical value, and using the least square method to fit to obtain the optimal values of k1, k2 and k3.

7. The method of claim 5, wherein, Also includes current saturation suppression step: the introduction of the diagonal weight matrix Ω, real-time monitoring of PCB coil current value and power amplifier state; When detecting the solution current of a coil exceeds the power amplifier saturation threshold, or the power amplifier reports the saturation state, the weight coefficient corresponding to the coil is exponentially multiplied, and the excitation current vector I is re-solved according to the formula I = Ω × G + × (G × Ω × G + ) - 1 × F, the current is limited in the saturation threshold range.

8. The method of claim 5, wherein, In step 4, the specific judgment logic of the coil dynamic gating is: if the magnetic field coupling strength of the PCB coil and the two-dimensional Halbach permanent magnet array (8) is greater than a set threshold value, it is determined that the coil is an "effective force coil" based on the pre-simulated magnetic field coupling curve, and the control system outputs a gating signal to the driving circuit of the coil.

9. The method of claim 5, wherein, In step 1, when establishing the high-precision electromagnetic force model (25), the fused pose information output by the data processing module is also combined, the data of the laser displacement sensor (4) and the capacitive sensor are fused by Kalman filtering to obtain, and the model parameters are real-time fine-tuned to further improve the electromagnetic force calculation accuracy.

Images