Numerical control rotary table real-time feedback control device based on multi-source data fusion

Abstract

The invention discloses a numerical control rotary table real-time feedback control device based on multi-source data fusion. The numerical control rotary table real-time feedback control device comprises a numerical control rotary table body, a multi-source data acquisition unit, a data fusion processing unit, an edge calculation module, a layered feedback control unit, a dynamic error compensation unit and a driving execution unit. The numerical control rotary table body adopts a cam roller gapless transmission structure; the multi-source data acquisition unit synchronously acquires position, temperature, vibration, torque and rotating speed full-dimensional data; the data fusion processing unit is used for realizing hardware-level data preprocessing; the edge calculation module is integrated with each unit onboard to realize local real-time calculation; the layered feedback control unit adopts a three-ring nested hardware control architecture; the dynamic error compensation unit is used for full-time dynamic compensation; all the units are electrically connected to form a full-hardware global closed-loop control system; the end-to-end response time of the device is less than or equal to 1ms, the positioning error is less than or equal to + / -0.5 arc second, the device is adaptive to various cam type numerical control rotary tables, and the localization of core parts of high-end numerical control equipment can be promoted.

Description

Technical Field

[0001] This invention relates to the field of precision control technology for CNC equipment, specifically to a real-time feedback control device for CNC rotary tables based on multi-source data fusion. It is particularly suitable for high-end CNC machine tools with cam-type four-axis rotary tables, five-axis rotary tables, and APC horizontal cam exchange tables that adopt a cam roller backlash-free transmission structure. It can meet the control requirements of high precision, high response speed, and high stability for the processing of complex irregular parts in aerospace, precision molds, medical devices, and other fields. Background Technology

[0002] As a core functional component of high-end CNC machine tools, the CNC rotary table directly determines the machining quality of complex curved surface parts through its positioning accuracy, repeatability, and operational stability, accounting for approximately 25% of the total cost of the CNC machine tool. Currently, over 90% of precision CNC rotary tables in China rely on imports, and existing domestic products and control devices suffer from numerous technical deficiencies, making it difficult to meet the demands of high-end machining.

[0003] First, existing CNC rotary table feedback control relies heavily on the position signal from a single encoder, failing to incorporate multi-source interference factors such as temperature, vibration, and load torque. During long-term continuous machining, thermal deformation of the cam drive mechanism, frictional vibration of the rotary table bearings, and fluctuations in cutting load can all lead to the accumulation of positioning errors. Single position feedback cannot detect these disturbances, resulting in inaccurate error compensation and difficulty in meeting the stringent precision requirements of the cam-roller backlash-free transmission structure.

[0004] Secondly, existing control devices lack efficient hardware-level fusion mechanisms for data processing. The sampling frequencies of different sensors vary significantly, resulting in asynchronous data in time and space. Furthermore, random noise and sudden interference are not effectively eliminated, leading to low reliability of feedback data and further impacting control accuracy. Simultaneously, data processing largely relies on the cloud or the machine tool's main controller, resulting in high transmission and processing delays and an inability to achieve millisecond-level real-time response, which is incompatible with the high-speed, precision transmission requirements of cam-type rotary tables.

[0005] Furthermore, traditional control structures often employ single-loop or dual-loop control, lacking a scientifically hierarchical hardware closed-loop control logic. Moreover, the control algorithms are mostly based on fixed parameters, unable to dynamically adjust according to load and speed, leading to overshoot during positioning and insufficient dynamic tracking accuracy. Error compensation modes are mostly static offline calibration or threshold-triggered intermittent compensation, generating compensation commands only when the error exceeds limits. This fails to adapt to dynamic errors caused by temperature, vibration, and load changes in real time, resulting in poor compensation effects and difficulty in achieving sub-arcsecond positioning. Simultaneously, existing technologies lack specific hardware compensation for the transmission backlash characteristics of cam-roller backlash-free transmission structures, failing to meet the high-precision control requirements of such turntables.

[0006] In addition, existing devices generally lack local real-time hardware computing capabilities and rely excessively on cloud or host computer transmission, which further exacerbates control delays; they lack an integrated real-time fault diagnosis mechanism, which cannot identify faults such as overload, overheating, and bearing wear in a timely manner, which can easily lead to equipment damage or workpiece scrap; lubrication and heat dissipation systems are mostly fixed modes and cannot be dynamically adjusted according to real-time temperature, which further aggravates thermal deformation errors and affects the service life and accuracy stability of the turntable.

[0007] In response to the shortcomings of existing technologies, the industry has also conducted relevant research: Existing technology 1, announcement number CN121477781A, discloses a CNC rotary table accuracy compensation method and system based on digital twins. It only collects three types of data: bearing temperature, servo motor current, and rotary table angular displacement. It cannot fully cover the core interference sources such as vibration and load torque during the operation of the CNC rotary table, making it difficult to achieve full-dimensional error compensation. Moreover, its core is a software simulation method based on a digital twin model, which relies on a host computer for simulation calculation. It has significant transmission and processing delays. It adopts threshold-triggered intermittent compensation and does not have a dedicated hardware closed-loop architecture for the cam roller transmission structure, so it cannot achieve continuous real-time compensation under all working conditions.

[0008] Existing technology 2, announcement number CN119794885B, discloses a method and device for thermal error compensation of CNC machine tools considering temperature and vibration coupling. Its technical solution is only for thermal error compensation of CNC machine tool electric spindle. The control object and application scenario are completely different from those of CNC rotary table. It does not provide any technical inspiration for transmission control and backlash compensation of cam-type CNC rotary table, and cannot provide technical reference for the full-link hardware closed-loop control of CNC rotary table.

[0009] To address the aforementioned issues, there is an urgent need to develop an integrated feedback control device that combines precise acquisition of multi-source data, efficient hardware-level fusion processing, hierarchical adaptive hardware control, real-time dynamic error compensation, local real-time calculation, fault diagnosis, and lubrication and heat dissipation linkage. This device would break through existing technological bottlenecks, improve the overall performance of CNC rotary tables, and support the domestic substitution of high-end CNC equipment. Summary of the Invention

[0010] The purpose of this invention is to provide a real-time feedback control device for CNC rotary tables based on multi-source data fusion, aiming to solve the following existing technical defects: single data acquisition dimension and incomplete perception of interference sources; reliance on software simulation / host computer processing without hardware-level fusion processing links, resulting in spatiotemporal asynchrony of multi-source data and high control delay; lack of a dedicated three-ring nested hardware control architecture for cam drive mechanisms, leading to low dynamic tracking accuracy and easy overshoot; the use of intermittent compensation mode without dedicated hardware compensation for cam transmission backlash, resulting in insufficient compensation accuracy; and the lack of an integrated hardware guarantee mechanism, limiting equipment reliability and service life. Simultaneously, this invention addresses the problems of unclear protection scope and insufficient inventive support caused by the inclusion of method features in the product claims in the original application, providing a real-time feedback control device for CNC rotary tables with prominent substantive features and significant progress.

[0011] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A real-time feedback control device for a CNC rotary table based on multi-source data fusion includes a CNC rotary table body, a multi-source data acquisition unit, a data fusion processing unit, an edge computing module, a hierarchical feedback control unit, a dynamic error compensation unit, and a drive execution unit. The CNC rotary table body includes a worktable, rotary table bearings, a cam drive mechanism and a mounting base. The cam drive mechanism adopts a cam roller backlash-free transmission structure. The multi-source data acquisition unit includes an absolute grating position sensor, a distributed PT100 temperature sensor, a triaxial MEMS vibration sensor, a non-contact torque sensor, and a high-resolution speed encoder. The absolute grating position sensor is fixedly installed at the rotation center between the worktable and the turntable bearing. The distributed PT100 temperature sensor is distributed at the core heat-generating points inside the mounting base, the cam drive mechanism, and the turntable bearing. The triaxial MEMS vibration sensor is fixed to the side wall of the mounting base near the cam drive mechanism. The non-contact torque sensor and the high-resolution speed encoder are coaxially connected in series at the input shaft end of the cam drive mechanism. The signal output terminals of each sensor are electrically connected to the signal input terminals of the data fusion processing unit. The data fusion processing unit is an FPGA preprocessing chip that integrates an adaptive Kalman filter circuit and a spatiotemporal synchronization alignment circuit. The input terminal of the adaptive Kalman filter circuit is connected to the signal output terminal of each sensor, and the input terminal of the spatiotemporal synchronization alignment circuit is connected to the output terminal of the adaptive Kalman filter circuit. The signal output terminal of the FPGA preprocessing chip is electrically connected to the signal input terminals of the edge computing module and the dynamic error compensation unit, respectively. The edge computing module adopts an industrial-grade edge computing chip. The industrial-grade edge computing chip, the FPGA preprocessing chip of the data fusion processing unit, and the hierarchical feedback control unit are integrated on the same industrial control board to form an onboard localized hardware processing link. The hierarchical feedback control unit is a three-ring nested industrial motion control chip integrating a position loop hardware control link, a speed loop hardware control link, and a torque loop hardware control link. The signal input terminal of the position loop hardware control link is electrically connected to the signal output terminal of the absolute grating position sensor, and the signal output terminal of the position loop hardware control link is electrically connected to the signal input terminal of the speed loop hardware control link. The signal input terminal of the speed loop hardware control link is electrically connected to the signal output terminal of the high-resolution speed encoder, and the signal output terminal of the speed loop hardware control link is electrically connected to the signal input terminal of the torque loop hardware control link. The signal input terminal of the torque loop hardware control link is electrically connected to the signal output terminal of the non-contact torque sensor, and the signal output terminal of the torque loop hardware control link is electrically connected to the first signal input terminal of the dynamic error compensation unit. The three-ring nested industrial motion control chip also integrates an adaptive PID parameter adjustment circuit corresponding one-to-one with each loop hardware control link. The dynamic error compensation unit is an industrial-grade MCU chip equipped with a fixed BP neural network prediction model. The second signal input terminal of the industrial-grade MCU chip is electrically connected to the signal output terminal of the FPGA preprocessing chip of the data fusion processing unit, and the signal output terminal of the industrial-grade MCU chip is electrically connected to the signal input terminal of the drive execution unit. The input layer of the fixed BP neural network prediction model is hardware-interfaced with the temperature, vibration, speed, and torque acquisition channels of the cam drive mechanism, and the output layer is hardware-interfaced with the pulse output circuits for thermal deformation compensation, vibration offset compensation, and cam transmission clearance compensation. The industrial-grade MCU chip integrates a hardware-level signal superposition circuit, and the two input terminals of the signal superposition circuit are electrically connected to the signal output terminal of the hierarchical feedback control unit and the compensation pulse output circuit, respectively. The drive execution unit includes a servo driver, a permanent magnet synchronous servo motor, and an electromagnetic braking module. The signal input terminal of the servo driver is electrically connected to the signal output terminal of the dynamic error compensation unit. The output shaft of the permanent magnet synchronous servo motor is coaxially connected to the input shaft of the cam drive mechanism. The electromagnetic braking module is coaxially assembled with the input shaft of the cam drive mechanism.

[0012] In a preferred embodiment, the absolute grating position sensor has an angular resolution ≤ 0.0001°; the distributed PT100 temperature sensor has 6-8 sampling points with a temperature measurement accuracy of ±0.1℃; the triaxial MEMS vibration sensor has a sampling frequency ≥ 1kHz; and the non-contact torque sensor has a response time ≤ 0.1ms.

[0013] In a preferred embodiment, the FPGA preprocessing chip of the data fusion processing unit has a multi-source data synchronization error of ≤1ms in its spatiotemporal synchronization alignment circuit.

[0014] In a preferred embodiment, the industrial-grade edge computing chip of the edge computing module has an end-to-end response time of ≤1ms for control commands.

[0015] In a preferred embodiment, the three-ring nested industrial motion control chip of the hierarchical feedback control unit has a position loop control bandwidth ≥200Hz, a speed loop response frequency ≥1kHz, and a dynamic tracking error ≤0.3 arcseconds.

[0016] In a preferred embodiment, the positioning error of the CNC rotary table is ≤ ±0.5 arcseconds after the industrial-grade MCU chip of the dynamic error compensation unit is compensated by a fixed BP neural network prediction model.

[0017] In a preferred embodiment, the system further includes a real-time fault diagnosis unit, which is a hardware diagnostic chip integrating a fixed fault feature threshold library. The signal input terminal of the hardware diagnostic chip is electrically connected to the signal output terminals of each sensor in the multi-source data acquisition unit, and the signal output terminal is electrically connected to the servo driver emergency stop control terminal of the drive execution unit. The fixed fault feature threshold library contains pre-stored vibration spectrum, temperature gradient, and torque fluctuation fault thresholds corresponding to the cam drive mechanism and the turntable bearing.

[0018] In a preferred embodiment, a lubrication and heat dissipation linkage unit is further included. The lubrication and heat dissipation linkage unit includes a variable frequency lubrication pump and a variable frequency cooling fan. The oil outlet of the variable frequency lubrication pump extends to the lubrication point of the cam drive mechanism and the turntable bearing. The variable frequency cooling fan is installed in the area of ​​the mounting base corresponding to the cam drive mechanism. The variable frequency control terminals of the variable frequency lubrication pump and the variable frequency cooling fan are electrically connected to the industrial-grade MCU chip of the dynamic error compensation unit.

[0019] In a preferred embodiment, the electromagnetic braking module of the drive execution unit has a braking positioning deviation ≤ 0.1 arcseconds and a braking torque ≥ 4600 N·m.

[0020] In a preferred embodiment, the CNC rotary table body is a cam-type four-axis rotary table, a five-axis rotary table, or an APC horizontal cam exchange table, and the maximum horizontal load of the worktable is ≥80kg.

[0021] Due to the application of the above technical solution, the beneficial effects of this application compared with the prior art are as follows: 1. Full hardware closed-loop architecture, completely breaking through the latency bottleneck of existing technologies: This invention adopts a full hardware onboard integrated architecture of "FPGA preprocessing chip + edge computing chip + three-ring nested motion control chip + solidified neural network MCU chip". All data processing, control calculation and compensation superposition are completed in the local hardware link. The end-to-end response time of control commands is ≤1ms, which completely solves the core defects of existing technology 1, which relies on upper computer software simulation and has high transmission and processing latency. It achieves millisecond-level real-time response and perfectly matches the high-speed and precision transmission requirements of cam-type turntables.

[0022] 2. Dedicated design for cam drive mechanisms, filling a gap in existing technology: This invention designs a dedicated sensor layout, a three-ring nested hardware control link, and a dedicated cam transmission backlash compensation circuit for the backlash-free transmission structure of cam rollers. It comprehensively covers all interference sources of the cam drive mechanism, including temperature, vibration, torque, position, and speed. This solves the industry pain points of existing technologies that do not have an adaptation design for cam transmission structures and have a single compensation dimension. After dedicated compensation, the turntable positioning error is ≤ ±0.5 arcseconds, and the dynamic tracking error is ≤0.3 arcseconds, achieving sub-arcsecond-level high-precision positioning.

[0023] 3. Hardware-level data fusion and continuous compensation for significantly improved performance: This invention achieves hardware-level preprocessing of multi-source data through the built-in filtering circuit and spatiotemporal synchronization alignment circuit of the FPGA chip, with a synchronization error of ≤1ms, solving the problems of spatiotemporal asynchrony and low reliability of multi-source data in the prior art; through the hardware-level signal superposition circuit of the MCU chip, it achieves continuous dynamic compensation under all working conditions, rather than the threshold-triggered intermittent compensation in the prior art 1, so that the turntable can process continuously for 72 hours without significant accuracy decay, and the accuracy retention is greatly improved.

[0024] 4. Integrated hardware support mechanism with outstanding reliability and adaptability: This invention integrates a hardware-level real-time fault diagnosis unit and a lubrication and heat dissipation linkage unit, with a fault identification accuracy rate of ≥99%, which can provide timely warnings and prevent equipment damage; by dynamically adjusting lubrication and heat dissipation parameters, the operating temperature fluctuation of the rotary table is ≤±1℃, and the overall service life is increased by more than 35%; the device can be directly mounted on cam-type four-axis and five-axis rotary tables and APC horizontal cam exchange tables, and can be seamlessly connected with existing CNC systems, breaking the technological monopoly of foreign brands and promoting the localization of core components of high-end CNC equipment. Attached Figure Description

[0025] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 This is an overall structural block diagram of a real-time feedback control device for a CNC rotary table based on multi-source data fusion according to the present invention. Figure 2 This is a schematic diagram of the CNC rotary table body structure of the present invention; Figure 3 This is a flowchart of the three-loop nested control of the hierarchical feedback control unit of the present invention; Figure 4 This is a schematic diagram of the BP neural network model structure of the dynamic error compensation unit of the present invention; Figure 5 This is a schematic diagram illustrating the workflow of a real-time feedback control device for a CNC rotary table based on multi-source data fusion according to the present invention. The components include: 1. CNC rotary table body; 11. Worktable; 12. Rotary table bearing; 13. Cam drive mechanism; 14. Mounting base; 2. Multi-source data acquisition unit; 21. Absolute grating position sensor; 22. Distributed PT100 temperature sensor; 23. Triaxial MEMS vibration sensor; 24. Non-contact torque sensor; 25. High-resolution speed encoder; 3. Data fusion processing unit; 31. Adaptive Kalman filter circuit; 32. Spatiotemporal synchronization alignment circuit; 4. Hierarchical feedback control unit; 41. Position loop hardware control link; 42. Speed ​​loop hardware control link; 43. Torque loop hardware control link; 5. Dynamic error compensation unit; 51. BP neural network prediction model; 6. Drive execution unit; 61. Servo driver; 62. Permanent magnet synchronous servo motor; 7. Edge computing module; 8. Real-time fault diagnosis unit; 81. Fixed fault characteristic threshold library; 9. Lubrication and heat dissipation linkage unit; 91. Variable frequency lubrication pump; 92. Variable frequency cooling fan. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] In this application, the terms "upper," "lower," "left," "right," "front," "rear," "top," "bottom," "inner," "outer," "middle," "vertical," "horizontal," "lateral," and "longitudinal" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are primarily for the purpose of better describing the invention and its embodiments, and are not intended to limit the indicated device, element, or component to having a specific orientation, or to be constructed and operated in a specific orientation.

[0030] Furthermore, in addition to indicating direction or positional relationship, some of the aforementioned terms may also have other meanings. For example, the term "above" may also be used in certain situations to indicate a dependency or connection. Those skilled in the art can understand the specific meaning of these terms in this invention based on the specific circumstances.

[0031] Furthermore, the terms "installation," "setup," "equipped with," "connection," "linking," and "socketing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral structure; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium, or an internal connection between two devices, components, or parts. Those skilled in the art can understand the specific meaning of these terms in this invention based on the specific circumstances.

[0032] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0033] Example 1 like Figure 1-5As shown, this embodiment provides a real-time feedback control device for CNC rotary tables based on multi-source data fusion, which is adapted to the machining of complex curved surfaces of aerospace engine blades and applied to high-precision machining scenarios of five-axis cam rotary tables. The specific structure is as follows: it includes a CNC rotary table body 1, a multi-source data acquisition unit 2, a data fusion processing unit 3, a hierarchical feedback control unit 4, a dynamic error compensation unit 5, a drive execution unit 6, an edge computing module 7, a real-time fault diagnosis unit 8, and a lubrication and heat dissipation linkage unit 9.

[0034] The CNC rotary table body 1 is a Jiangsu Gutian AC five-axis cam rotary table, which includes a worktable 11, a rotary table bearing 12, a cam drive mechanism 13 and a mounting base 14. The cam drive mechanism 13 adopts a cam roller backlash-free transmission structure. The worktable 11 has a diameter of 250mm and a maximum horizontal load of 100kg.

[0035] The multi-source data acquisition unit 2 includes an absolute grating position sensor 21, a distributed PT100 temperature sensor 22, a triaxial MEMS vibration sensor 23, a non-contact torque sensor 24, and a high-resolution speed encoder 25. Wherein: The absolute grating position sensor 21 adopts RENISHAW RGH25Z with an angular resolution of 0.0001°. It is fixedly installed at the rotation center between the worktable 11 and the turntable bearing 12. The signal output end is electrically connected to the data fusion processing unit 3 and the hierarchical feedback control unit 4. The distributed PT100 temperature sensor 22 has Class A accuracy and a total of 8 sampling points. It is located at the core heating points inside the mounting base 14, cam drive mechanism 13, and turntable bearing 12. The temperature measurement accuracy is ±0.1℃. The signal output terminal is electrically connected to the data fusion processing unit 3. The triaxial MEMS vibration sensor 23 uses an ADI ADXL355 with a sampling frequency of 1kHz. It is fixed to the side wall of the mounting base 14 near the cam drive mechanism 13, and its signal output terminal is electrically connected to the data fusion processing unit 3.

[0036] The non-contact torque sensor 24 adopts HJN-105, with a range of 0-5000 N·m and a response time of ≤0.1 ms. It is coaxially connected in series with the high-resolution speed encoder 25 (Pepperl+Fuchs RVI58N, 1024 lines) at the input shaft end of the cam drive mechanism 13. The signal output ends of both are electrically connected to the data fusion processing unit 3 and the hierarchical feedback control unit 4.

[0037] The data fusion processing unit 3 is a Xilinx Artix-7 FPGA preprocessing chip, which integrates an adaptive Kalman filter circuit 31 and a spatiotemporal synchronization alignment circuit 32. The input of the adaptive Kalman filter circuit 31 is connected to the signal output of each sensor to remove random noise and sudden interference caused by cutting impact in the data at the hardware level. The input of the spatiotemporal synchronization alignment circuit 32 is connected to the output of the adaptive Kalman filter circuit 31 to unify multi-source data with different sampling frequencies to the same time axis, with a data synchronization error ≤1ms. The signal output of the FPGA preprocessing chip is electrically connected to the edge computing module 7 and the dynamic error compensation unit 5, respectively.

[0038] The edge computing module 7 uses the NVIDIA Jetson Nano industrial-grade edge computing chip with a computing power of 472 GFLOPS. It is integrated with the FPGA preprocessing chip and the hierarchical feedback control unit 4 on the same industrial control board to form an onboard localized hardware processing link, with an end-to-end response time of control commands ≤1ms.

[0039] The hierarchical feedback control unit 4 is a three-ring nested industrial motion control chip integrating a position loop hardware control link 41, a speed loop hardware control link 42, and a torque loop hardware control link 43. The chip model is STM32H743ZI. Wherein: The signal input terminal of the position loop hardware control link 41 is electrically connected to the signal output terminal of the absolute grating position sensor 21, and the output terminal is electrically connected to the input terminal of the speed loop hardware control link 42, with a control bandwidth of 220Hz. The signal input terminal of the speed loop hardware control link 42 is electrically connected to the signal output terminal of the high-resolution speed encoder 25, and the output terminal is electrically connected to the input terminal of the torque loop hardware control link 43, with a response frequency of 1.2kHz. The signal input terminal of the torque loop hardware control link 43 is electrically connected to the signal output terminal of the non-contact torque sensor 24, and the output terminal is electrically connected to the first signal input terminal of the dynamic error compensation unit 5. The dynamic tracking error is ≤0.3 arcseconds. The industrial motion control chip also integrates an adaptive PID parameter adjustment circuit that corresponds one-to-one with each hardware control link, which can adjust the proportional, integral, and derivative control parameters in real time according to the load and speed at the hardware level.

[0040] The dynamic error compensation unit 5 is an STM32H750VB industrial-grade MCU chip, which integrates a fixed BP neural network prediction model 51 and a hardware-level signal superposition circuit. The input layer of the fixed BP neural network prediction model 51 is hardware-interfaced with the temperature, vibration, speed, and torque acquisition channels of the cam drive mechanism 13, and the output layer is hardware-interfaced with the pulse output circuits for thermal deformation compensation, vibration offset compensation, and cam transmission clearance compensation. The two input terminals of the hardware-level signal superposition circuit are electrically connected to the signal output terminal of the hierarchical feedback control unit 4 and the compensation pulse output circuit, respectively, to realize the hardware-level real-time superposition of the error compensation value and the basic control command, and output the final control command. After compensation, the positioning error of the CNC rotary table is ≤ ±0.5 arcseconds.

[0041] The drive execution unit 6 includes a servo driver 61, a permanent magnet synchronous servo motor 62, and an electromagnetic braking module. The servo driver 61 is a Panasonic MDDLN55SE, and its signal input terminal is electrically connected to the signal output terminal of the dynamic error compensation unit 5. The permanent magnet synchronous servo motor 62 is a 5.5kW permanent magnet synchronous motor, and its output shaft is coaxially connected to the input shaft of the cam drive mechanism 13. The electromagnetic braking module is coaxially assembled with the input shaft of the cam drive mechanism 13, with a braking positioning deviation ≤0.1 arcseconds and a braking torque ≥4600N·m.

[0042] The real-time fault diagnosis unit 8 is a hardware diagnostic chip that integrates a fixed fault feature threshold library 81. The hardware diagnostic chip adopts STM32F103. Its signal input terminal is electrically connected to the signal output terminals of each sensor in the multi-source data acquisition unit 2, and its signal output terminal is electrically connected to the emergency stop control terminal of the servo driver 61. The fixed fault feature threshold library 81 pre-stores the vibration spectrum, temperature gradient, and torque fluctuation fault thresholds corresponding to the cam drive mechanism 13 and the turntable bearing 12, with a fault identification accuracy of ≥99%.

[0043] The lubrication and heat dissipation linkage unit 9 includes a variable frequency lubrication pump 91 and a variable frequency cooling fan 92. The oil outlet of the variable frequency lubrication pump 91 extends to the lubrication point of the cam drive mechanism 13 and the turntable bearing 12, and the oil supply flow rate can be adjusted from 0 to 5 L / min. The variable frequency cooling fan 92 is installed in the area of ​​the mounting base 14 corresponding to the cam drive mechanism 13, and the power can be adjusted from 30% to 100%. The variable frequency control terminals of the variable frequency lubrication pump 91 and the variable frequency cooling fan 92 are electrically connected to the industrial-grade MCU chip, which can automatically adjust the operating parameters according to the real-time temperature to maintain the working temperature fluctuation of the CNC turntable ≤ ±1℃.

[0044] The performance test results of this embodiment are as follows: positioning accuracy ±0.45 arcseconds, repeatability 3.8 arcseconds, end-to-end response time 0.8ms, operating temperature fluctuation ±0.8℃, and no significant accuracy decay after 72 hours of continuous processing, fully meeting the high-end processing requirements of complex aerospace parts.

[0045] Example 2 This embodiment provides a real-time feedback control device for CNC rotary tables based on multi-source data fusion, adapted for high-speed milling of precision molds, and applied to high-speed machining scenarios of four-axis cam rotary tables. The core difference from Embodiment 1 is: The CNC rotary table body 1 is a Jiangsu Gutian NC four-axis cam rotary table, the worktable 11 has a diameter of 200mm, and the maximum horizontal load is 80kg; The absolute grating position sensor 21 uses RENISHAW RGH34Z with an angular resolution of 0.00008°; the triaxial MEMS vibration sensor 23 uses ADI ADXL375 with a sampling frequency of 2kHz; and the non-contact torque sensor 24 uses the HJN-106 high dynamic model with a response time ≤0.1ms. The data fusion processing unit 3 uses an STM32H750VB microcontroller with a data processing latency of ≤0.3ms; the hierarchical feedback control unit 4 uses a Xilinx Kintex-7 FPGA with a position loop bandwidth of 300Hz, a velocity loop response frequency of 2kHz, and a dynamic tracking error of ≤0.2 arcseconds. The edge computing module 7 uses NVIDIA Jetson Xavier NX with a computing power of 21 TOPS; the fixed BP neural network prediction model 51 of the dynamic error compensation unit 5 is optimized for high-speed operation, and the prediction cycle is shortened to 0.3ms. The drive unit 6 uses a Panasonic MDDLN65SE driver and a 7.5kW high-speed permanent magnet synchronous motor with a maximum operating speed of 500rpm.

[0046] The performance test results of this embodiment are as follows: positioning accuracy ±0.38 arcseconds, repeatability 3.2 arcseconds, response time 0.4ms, dynamic tracking error 0.18 arcseconds, and accuracy decay ≤0.03 arcseconds after 48 hours of continuous high-speed operation, perfectly meeting the high-speed and high-precision machining requirements of precision molds.

[0047] The remaining structure is the same as in Embodiment 1, and will not be described again here.

[0048] Example 3 This embodiment provides a real-time feedback control device for CNC rotary tables based on multi-source data fusion, adapted to multi-station batch processing of automotive parts, and applied to multi-station synchronous control scenarios of APC horizontal cam exchange tables. The core difference from Embodiment 1 is: The CNC rotary table body 1 is a Jiangsu Gutian APC horizontal cam exchange table with a double worktable design 11. The maximum load of a single table is 150kg, and it adopts a cam roller synchronous transmission structure. The multi-source data acquisition unit 2 is equipped with dual absolute grating position sensors 21 to synchronously acquire the positions of the two worktables; it is equipped with 10 distributed PT100 temperature sensors 22 to cover the two worktables and the exchange mechanism; and it is equipped with dual triaxial MEMS vibration sensors 23 and dual torque sensors. The data fusion processing unit 3 adopts dual STM32H743ZI master-slave collaborative microcontrollers, with a dual-workstation data synchronization error of ≤0.8ms; the hierarchical feedback control unit 4 has a built-in dual-station synchronization control algorithm, with a dual-workstation synchronization error of ≤1 arcsecond; The dynamic error compensation unit 5 is configured with a dual-fixed BP neural network prediction model 51, with the input of a synchronization deviation variable. After compensation, the synchronization error of the two worktables is ≤0.5 arcseconds. The drive unit 6 uses a Panasonic MDDLN85SE dual driver, an 11kW permanent magnet synchronous motor, and a 6000N·m dual independent braking electromagnetic braking module.

[0049] The performance test results of this embodiment are as follows: single worktable positioning accuracy ±0.42 arcseconds, dual worktable synchronization error 0.45 arcseconds, table change time 1.8s, and synchronization accuracy decay ≤0.1 arcseconds after 1000 consecutive table changes, meeting the high efficiency and high precision requirements of batch processing of automotive parts.

[0050] The remaining structure is the same as in Embodiment 1, and will not be described again here.

[0051] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time feedback control device for a CNC rotary table based on multi-source data fusion, characterized in that, It includes a CNC rotary table body, a multi-source data acquisition unit, a data fusion processing unit, an edge computing module, a hierarchical feedback control unit, a dynamic error compensation unit, and a drive execution unit; The CNC rotary table body includes a worktable, rotary table bearings, a cam drive mechanism and a mounting base. The cam drive mechanism adopts a cam roller backlash-free transmission structure. The multi-source data acquisition unit includes an absolute grating position sensor, a distributed PT100 temperature sensor, a triaxial MEMS vibration sensor, a non-contact torque sensor, and a high-resolution speed encoder. The absolute grating position sensor is fixedly installed at the rotation center between the worktable and the turntable bearing. The distributed PT100 temperature sensor is distributed at the core heat-generating points inside the mounting base, the cam drive mechanism, and the turntable bearing. The triaxial MEMS vibration sensor is fixed to the side wall of the mounting base near the cam drive mechanism. The non-contact torque sensor and the high-resolution speed encoder are coaxially connected in series at the input shaft end of the cam drive mechanism. The signal output terminals of each sensor are electrically connected to the signal input terminals of the data fusion processing unit. The data fusion processing unit is an FPGA preprocessing chip that integrates an adaptive Kalman filter circuit and a spatiotemporal synchronization alignment circuit. The input terminal of the adaptive Kalman filter circuit is connected to the signal output terminal of each sensor, and the input terminal of the spatiotemporal synchronization alignment circuit is connected to the output terminal of the adaptive Kalman filter circuit. The signal output terminal of the FPGA preprocessing chip is electrically connected to the signal input terminals of the edge computing module and the dynamic error compensation unit, respectively. The edge computing module adopts an industrial-grade edge computing chip. The industrial-grade edge computing chip, the FPGA preprocessing chip of the data fusion processing unit, and the hierarchical feedback control unit are integrated on the same industrial control board to form an onboard localized hardware processing link. The hierarchical feedback control unit is a three-ring nested industrial motion control chip integrating a position loop hardware control link, a speed loop hardware control link, and a torque loop hardware control link. The signal input terminal of the position loop hardware control link is electrically connected to the signal output terminal of the absolute grating position sensor, and the signal output terminal of the position loop hardware control link is electrically connected to the signal input terminal of the speed loop hardware control link. The signal input terminal of the speed loop hardware control link is electrically connected to the signal output terminal of the high-resolution speed encoder, and the signal output terminal of the speed loop hardware control link is electrically connected to the signal input terminal of the torque loop hardware control link. The signal input terminal of the torque loop hardware control link is electrically connected to the signal output terminal of the non-contact torque sensor, and the signal output terminal of the torque loop hardware control link is electrically connected to the first signal input terminal of the dynamic error compensation unit. The three-ring nested industrial motion control chip also integrates an adaptive PID parameter adjustment circuit corresponding one-to-one with each loop hardware control link. The dynamic error compensation unit is an industrial-grade MCU chip equipped with a fixed BP neural network prediction model. The second signal input terminal of the industrial-grade MCU chip is electrically connected to the signal output terminal of the FPGA preprocessing chip of the data fusion processing unit, and the signal output terminal of the industrial-grade MCU chip is electrically connected to the signal input terminal of the drive execution unit. The input layer of the fixed BP neural network prediction model is hardware-interfaced with the temperature, vibration, speed, and torque acquisition channels of the cam drive mechanism, and the output layer is hardware-interfaced with the pulse output circuits for thermal deformation compensation, vibration offset compensation, and cam transmission clearance compensation. The industrial-grade MCU chip integrates a hardware-level signal superposition circuit, and the two input terminals of the signal superposition circuit are electrically connected to the signal output terminal of the hierarchical feedback control unit and the compensation pulse output circuit, respectively. The drive execution unit includes a servo driver, a permanent magnet synchronous servo motor, and an electromagnetic braking module. The signal input terminal of the servo driver is electrically connected to the signal output terminal of the dynamic error compensation unit. The output shaft of the permanent magnet synchronous servo motor is coaxially connected to the input shaft of the cam drive mechanism. The electromagnetic braking module is coaxially assembled with the input shaft of the cam drive mechanism.

2. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The absolute grating position sensor has an angular resolution ≤0.0001°; the distributed PT100 temperature sensor has 6-8 sampling points with a temperature measurement accuracy of ±0.1℃; the triaxial MEMS vibration sensor has a sampling frequency ≥1kHz; and the non-contact torque sensor has a response time ≤0.1ms.

3. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The FPGA preprocessing chip of the data fusion processing unit has a multi-source data synchronization error of ≤1ms in its spatiotemporal synchronization alignment circuit.

4. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The industrial-grade edge computing chip of the edge computing module has an end-to-end response time of ≤1ms for control commands.

5. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The three-ring nested industrial motion control chip of the hierarchical feedback control unit has a position loop control bandwidth ≥200Hz, a speed loop response frequency ≥1kHz, and a dynamic tracking error ≤0.3 arcseconds.

6. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The industrial-grade MCU chip of the dynamic error compensation unit, after being compensated by a fixed BP neural network prediction model, has a positioning error of ≤±0.5 arcseconds for the CNC rotary table.

7. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, It also includes a real-time fault diagnosis unit, which is a hardware diagnostic chip that integrates a fixed fault feature threshold library. The signal input terminal of the hardware diagnostic chip is electrically connected to the signal output terminals of each sensor of the multi-source data acquisition unit, and the signal output terminal is electrically connected to the servo driver emergency stop control terminal of the drive execution unit. The fixed fault feature threshold library pre-stores the vibration spectrum, temperature gradient, and torque fluctuation fault thresholds corresponding to the cam drive mechanism and the turntable bearing.

8. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 2, characterized in that, It also includes a lubrication and heat dissipation linkage unit, which includes a variable frequency lubrication pump and a variable frequency cooling fan. The oil outlet of the variable frequency lubrication pump extends to the lubrication point of the cam drive mechanism and the turntable bearing. The variable frequency cooling fan is installed in the area of ​​the mounting base corresponding to the cam drive mechanism. The variable frequency control terminals of the variable frequency lubrication pump and the variable frequency cooling fan are electrically connected to the industrial-grade MCU chip of the dynamic error compensation unit.

9. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The electromagnetic braking module of the drive execution unit has a braking positioning deviation of ≤0.1 arcseconds and a braking torque of ≥4600 N·m.

10. The real-time feedback control device for a CNC rotary table based on multi-source data fusion according to claim 1, characterized in that, The CNC rotary table body is a cam-type four-axis rotary table, a five-axis rotary table, or an APC horizontal cam exchange table, and the maximum horizontal load of the worktable is ≥80kg.

Images