Servo position control method and system applied to engraving and milling machine

Abstract

The invention relates to the technical field of program control, in particular to a servo position control method and system applied to an engraving and milling machine. The method comprises the following steps that real-time temperature data and real-time load data of lead screws of all shafts in the engraving and milling machine are obtained; collecting multi-node temperature distribution data of the engraving and milling machine; obtaining cutting force data and resonant frequency characteristic values in the engraving and milling machine; constructing a screw rod reverse clearance prediction model based on the real-time temperature data and the real-time load data of the screw rod, and generating a clearance compensation value; performing pre-compensation control before the servo position direction of each axis is changed based on the clearance compensation value to obtain a correction position instruction; and performing thermal deformation simulation on the engraving and milling machine based on the multi-node temperature distribution data to generate thermal deformation compensation amounts of all shafts. Through multi-source data fusion and intelligent control, comprehensive optimization of the servo position system of the engraving and milling machine is achieved, the machining precision, dynamic response and stability are effectively improved, and efficient and reliable operation of high-speed and high-precision machining is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of program control technology, and in particular to a servo position control method and system applied to an engraving and milling machine. Background Art

[0002] Milling machines are widely used in mold manufacturing, precision parts processing, and other fields. Servo position control, a key technology influencing machining accuracy and dynamic performance, has evolved from traditional open-loop control and closed-loop position control to the current multi-dimensional control strategy with integrated compensation and intelligent control. Current milling machine servo control systems often use PID control based on position feedback, combined with certain feedforward compensation methods, to achieve precise control of each axis' position.

[0003] Although the existing servo position control method can achieve basic high-precision positioning, it still faces many challenges in the high-speed and high-precision machining process of engraving and milling machines. First, the reverse clearance changes caused by the thermal expansion of the screw and the load change are difficult to compensate accurately in real time, resulting in the accumulation of position command errors. Secondly, the nonlinear thermal deformation of the machine tool structure in a complex thermal environment is difficult to model and compensate, which affects the machining stability. Thirdly, traditional feed control does not take into account the real-time cutting force changes and resonance characteristics, making it difficult to achieve dynamic adaptive adjustment, which reduces the trajectory smoothness and machining efficiency. In addition, the lag in the response of the servo control loop to the actual feedback error also limits the further improvement of the overall control accuracy. Summary of the Invention

[0004] Based on this, it is necessary for the present invention to provide a servo position control method and system applied to an engraving and milling machine to solve at least one of the above technical problems.

[0005] To achieve the above object, a servo position control method applied to an engraving and milling machine includes the following steps:

[0006] Step S1: obtaining the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collecting the multi-node temperature distribution data of the engraving and milling machine; obtaining the cutting force data and resonance frequency characteristic value in the engraving and milling machine;

[0007] Step S2: constructing a screw rod backlash prediction model based on the screw rod real-time temperature data and real-time load data, and generating a backlash compensation value; performing pre-compensation control before the servo position direction of each axis is changed based on the backlash compensation value to obtain a corrected position command;

[0008] Step S3: performing thermal deformation simulation on the engraving and milling machine based on the multi-node temperature distribution data to generate thermal deformation compensation for each axis;

[0009] Step S4: adjusting the feed speed of the engraving and milling machine based on the cutting force data to obtain an optimized feed rate coefficient;

[0010] Step S5: Generate a comprehensive control instruction based on the corrected position instruction, the thermal deformation compensation amount of each axis, and the optimized feed rate coefficient; construct a machining trajectory of the engraving and milling machine based on the optimized feed rate coefficient and the resonance frequency characteristic value, and perform optimized interpolation to obtain smoothed trajectory parameters; coordinately drive and control the servo motors of each axis based on the comprehensive control instruction and the smoothed trajectory parameters to obtain actual position feedback data of each axis;

[0011] Step S6: Evaluate the control accuracy of the actual position feedback data of each axis to obtain position error data; update the comprehensive control instructions based on the position error data, and perform dual-drive coordinated control on the engraving and milling machine.

[0012] Through multi-source data fusion and intelligent modeling, this invention achieves comprehensive optimization of the servo position control system for milling machines, significantly improving their machining accuracy and dynamic response capabilities. First, by incorporating screw temperature, load, and multi-node temperature data from the entire machine, it effectively addresses the issues of insufficient compensation for gap errors and structural deformation caused by thermal expansion, load variations, and nonlinear thermal deformation, enhancing the system's robustness to thermal environment fluctuations. Second, by combining cutting force and resonance characteristic information, it enables dynamic adaptive adjustment of feed rate and trajectory smoothing interpolation, significantly improving trajectory smoothness and machining efficiency during milling and reducing trajectory deviations and surface defects caused by vibration. Furthermore, by performing real-time accuracy assessment and closed-loop error updates based on the actual position feedback of each axis, the system further enhances the response speed and accuracy of position control, ensuring stable operation under complex machining conditions. Finally, through a dual-drive coordinated control approach, the limitations of traditional control strategies in large dynamic loads and high-precision synchronous drive are effectively overcome, achieving higher-precision and more stable servo control performance, and overall improving the comprehensive performance and intelligence level of milling machines for high-speed, high-precision machining tasks.

[0013] Preferably, the present invention further provides a servo position control system for an engraving and milling machine, which is used to execute the above-mentioned servo position control method for an engraving and milling machine. The servo position control system for an engraving and milling machine comprises:

[0014] Multi-source data acquisition module, used to obtain the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collect the multi-node temperature distribution data of the engraving and milling machine; obtain the cutting force data and resonance frequency characteristic value of the engraving and milling machine;

[0015] The intelligent backlash prediction and compensation module is used to build a screw backlash prediction model based on the screw's real-time temperature and load data, and generate backlash compensation values. Based on the backlash compensation values, pre-compensation control is performed before the servo position and direction of each axis are changed to obtain a corrected position command.

[0016] The digital twin thermal deformation simulation module is used to simulate the thermal deformation of the engraving and milling machine based on multi-node temperature distribution data and generate thermal deformation compensation for each axis;

[0017] Adaptive feed adjustment module, used to adjust the feed speed of the engraving and milling machine based on cutting force data to obtain the optimized feed rate coefficient;

[0018] The multi-axis coordinated control module is used to generate comprehensive control instructions based on the corrected position instructions, thermal deformation compensation of each axis, and optimized feed rate coefficients; construct the machining trajectory of the engraving and milling machine based on the optimized feed rate coefficients and resonance frequency characteristic values, and perform optimized interpolation to obtain smoothed trajectory parameters; based on the comprehensive control instructions and smoothed trajectory parameters, the servo motors of each axis are coordinated and controlled to obtain actual position feedback data of each axis;

[0019] The precision assessment feedback optimization module is used to evaluate the control precision of the actual position feedback data of each axis to obtain the position error data; based on the position error data, it updates the comprehensive control instructions and performs dual-drive coordinated control of the engraving and milling machine.

[0020] Through modular design, the present invention achieves intelligent optimization of the entire servo control process of the milling machine, achieving significant comprehensive improvement effects. The integrated acquisition of multi-dimensional data provides a comprehensive, real-time perception foundation for the control system, making control instructions more environmentally adaptable and decision-making accurate. Dynamic predictive compensation of backlash improves the foresight and accuracy of displacement control, effectively reducing trajectory deviation caused by mechanical errors. Thermal deformation simulation and compensation enhance the system's response to structural changes caused by temperature rise, significantly improving machining stability under complex thermal fields. Through adaptive feed rate adjustment, the system can respond to cutting load fluctuations in real time, ensuring the smoothness of the machining trajectory and the reliability of equipment operation. Trajectory interpolation optimization and multi-axis coordinated drive control improve the consistency of servo response and reduce the impact of vibration interference on machining quality. Furthermore, through precision assessment and feedback closed-loop correction mechanisms, the system has self-learning and continuous optimization capabilities, making servo control more refined and intelligent. Finally, through the coordinated control of the main drive and preload motor, the dynamic stability and synchronization performance of the control system in high-precision and high-speed machining scenarios are further enhanced, comprehensively improving the machining accuracy, response speed, and operating efficiency of the milling machine. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0022] Figure 1 A schematic flow chart of the steps of a servo position control method for an engraving and milling machine according to the present invention;

[0023] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0024] Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0025] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0026] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0027] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0028] To achieve this, please refer to Figures 1 to 3 The present invention provides a servo position control method for a milling machine, the method comprising the following steps:

[0029] Step S1: obtaining the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collecting the multi-node temperature distribution data of the engraving and milling machine; obtaining the cutting force data and resonance frequency characteristic value in the engraving and milling machine;

[0030] Step S2: constructing a screw rod backlash prediction model based on the screw rod real-time temperature data and real-time load data, and generating a backlash compensation value; performing pre-compensation control before the servo position direction of each axis is changed based on the backlash compensation value to obtain a corrected position command;

[0031] Step S3: performing thermal deformation simulation on the engraving and milling machine based on the multi-node temperature distribution data to generate thermal deformation compensation for each axis;

[0032] Step S4: adjusting the feed speed of the engraving and milling machine based on the cutting force data to obtain an optimized feed rate coefficient;

[0033] Step S5: Generate a comprehensive control instruction based on the corrected position instruction, the thermal deformation compensation amount of each axis, and the optimized feed rate coefficient; construct a machining trajectory of the engraving and milling machine based on the optimized feed rate coefficient and the resonance frequency characteristic value, and perform optimized interpolation to obtain smoothed trajectory parameters; coordinately drive and control the servo motors of each axis based on the comprehensive control instruction and the smoothed trajectory parameters to obtain actual position feedback data of each axis;

[0034] Step S6: Evaluate the control accuracy of the actual position feedback data of each axis to obtain position error data; update the comprehensive control instructions based on the position error data, and perform dual-drive coordinated control on the engraving and milling machine.

[0035] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a flow chart of a servo position control method for a milling machine according to the present invention. In this embodiment, the servo position control method for a milling machine includes the following steps:

[0036] Step S1: obtaining the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collecting the multi-node temperature distribution data of the engraving and milling machine; obtaining the cutting force data and resonance frequency characteristic value in the engraving and milling machine;

[0037] In this embodiment of the present invention, thermocouple temperature sensors with an accuracy of no less than ±0.1°C are fixedly mounted on the ball screws of the X-, Y-, and Z-axes. Temperature data is acquired through high-frequency (no less than 1Hz) continuous sampling. Data sampling is achieved through an embedded data acquisition card connected to the industrial control system. Axial load on each axis is collected by connecting a strain gauge current sensor with a range of 0–5000N and a sensitivity of 1mV / V in series to each servo driver's current path. This sensor detects changes in the servo motor's drive current in real time and converts it into load data. Load conversion uses the standard motor torque constant and lead formula: load = (current × motor torque constant) ÷ lead. During the multi-node temperature distribution data collection process for the engraving and milling machine bed, column, spindle box, and worktable, no fewer than eight semiconductor temperature sensors (with an accuracy better than ±0.2°C) are evenly distributed across the four structural components. Each sensor transmits data to the central processing unit via the CAN bus, and the sensor's spatial coordinates are simultaneously recorded to construct a temperature distribution map. The distribution points should cover the main heat transfer paths and structural connection points, such as the intersection of the spindle box and column, the front and rear end surfaces of the worktable, the center of the bed base, and the symmetrical points on both sides. Cutting force data is collected using a three-axis piezoelectric force sensor (with a sensitivity of no less than 5pC / N) mounted on the tool holder. The cutting forces in the X / Y / Z directions are acquired at a 0.5ms cycle, and the corresponding spindle speed and feed rate are simultaneously recorded and superimposed. The data is input into the signal conditioning module after isolation and amplification, and is sampled and stored in parallel by a high-speed acquisition card. To obtain the resonance frequency characteristic value, it is necessary to attach three-axis vibration acceleration sensors (frequency response range 0.5–5000 Hz) to the center axis of the machine tool spindle and the end of each guide rail. Then, the machine tool's no-load speed-changing motion program is run to stimulate the resonance of the entire machine structure at different frequencies. The vibration acceleration is fast Fourier transformed using the spectrum analysis module, and the frequency corresponding to the peak amplitude in the frequency domain is recorded. The frequency point with the maximum amplitude is selected as the resonance frequency characteristic value and written into the control parameter buffer.

[0038] Step S2: constructing a screw rod backlash prediction model based on the screw rod real-time temperature data and real-time load data, and generating a backlash compensation value; performing pre-compensation control before the servo position direction of each axis is changed based on the backlash compensation value to obtain a corrected position command;

[0039] In this embodiment of the present invention, the real-time temperature data of each ball screw axis collected in step S1 is first transmitted to a filtering processing unit at a sampling frequency of 1 Hz. A weighted average method is used to eliminate abnormal fluctuations. The temperature data at five consecutive time points is smoothed by a linear combination with weighting coefficients of 0.1, 0.2, 0.4, 0.2, and 0.1. Before entering subsequent processing, the real-time load data is scaled to a value between 0 and 1 based on the full-scale value and actual measured value of each current sensor to obtain normalized load data. Thermal expansion is then calculated based on the ball screw material and its structural parameters using the formula: thermal expansion = linear expansion coefficient × temperature change × screw length. This thermal expansion factor is then constructed by combining this thermal expansion with the normalized load data. Next, the bidirectional displacement detection module measures the minimum starting displacement difference between the ball screw and nut pairs in forward and reverse driving directions, defining this as the transmission clearance reference value. An axial load is applied using a static pressure loading platform with a loading force of 1000N. The relative displacement between the screw and nut pair is measured, and the load deformation value is calculated based on this displacement. The load deformation coefficient is obtained using the formula: deformation coefficient = load deformation / loading force. The thermal expansion factor and the load deformation coefficient are combined and input into the clearance prediction processing module. A linear weighted relationship is used to generate a predicted value for the ball screw reverse clearance. This relationship assigns an empirical weight of 0.6 to the thermal expansion factor and 0.4 to the load deformation coefficient. The prediction formula is: clearance = 0.6 × thermal expansion factor + 0.4 × load deformation coefficient. The predicted value is combined with the transmission direction detection signal (obtained from the encoder to detect the motor rotation direction) to determine whether the motion is in reverse. If the direction is reverse, the original predicted value is used as the reverse compensation value. If the direction is forward, the reverse compensation value is set to 0. The forward compensation value is set to 0.01mm using an empirical value, forming a complete set of directional compensation values. The forward and reverse compensation values ​​are vector-superimposed before the command position signal is output, and the corrected position command is output.

[0040] Step S3: performing thermal deformation simulation on the engraving and milling machine based on the multi-node temperature distribution data to generate thermal deformation compensation for each axis;

[0041] In the embodiment of the present invention, the temperature data of 32 measuring points distributed on the engraving and milling machine bed, column, spindle box and workbench collected in step S1 are first imported into the three-dimensional coordinate mapping unit, and the physical space coordinates of each measuring point in the machine tool structure are combined to construct a three-dimensional temperature field using interpolation operation, wherein the interpolation method adopts a trilinear interpolation method, and the grid accuracy is set such that the side length of each cubic unit does not exceed 50 mm to generate a three-dimensional temperature field model; then, based on the temperature field model, the thermal deformation simulation of the guide rail mounting surface of the engraving and milling machine bed is performed, and the calculation method is as follows: the bed structure is divided into multiple finite size units, and the material of each unit is set to HT250 gray cast iron. The thermal strain accumulation formula ΔL=α·ΔT·L driven by temperature difference is used to calculate the X-axis respectively. The thermal expansion of the guide rail reference line in the direction is superimposed to form the thermal deformation of the bed; then, based on the temperature gradient difference in the connection area between the spindle box and the column, the change in the column inclination angle in the Z-axis direction is calculated according to the linear thermal elongation method between the nodes, and then the thermal offset displacement in the Y-axis direction is derived by multiplying the inclination angle by the spindle center height to obtain the column thermal inclination deformation; on the spindle centerline, the front and rear temperature difference ΔT of the bearing seat in the spindle box is taken, combined with the bearing seat spacing, according to the thermal displacement calculation formula Δx=α·ΔT·L, the spindle thermal displacement is output; for the thermal deformation simulation of the workbench, a bidirectional linear interpolation temperature distribution surface is constructed with the four corner measuring points of the surface as boundary conditions, and the table warpage is calculated by the isothermal line centroid method and converted into a planar thermal deformation in the Z-axis direction. Integrating the above four types of structural deformation data, the three-axis thermal deformation is calculated based on the following compensation relationship: the X-axis thermal deformation compensation is determined by the sum of the bed thermal deformation and the spindle thermal displacement, the Y-axis thermal deformation compensation is determined by the superposition of the bed thermal deformation, the column thermal tilt deformation and the spindle thermal displacement, and the Z-axis thermal deformation compensation is determined by the combination of the column thermal tilt deformation and the spindle thermal displacement. The three-axis compensation units are unified in millimeters, with three decimal places retained and written into the control parameter buffer.

[0042] Step S4: adjusting the feed speed of the engraving and milling machine based on the cutting force data to obtain an optimized feed rate coefficient;

[0043] In this embodiment of the present invention, the cutting force data acquired in step S1 is first input into a real-time filtering unit at 0.5ms intervals. A median filter with a sliding window of 5 points is used to eliminate transient pulse interference, generating smoothed cutting force data. This smoothed cutting force data is then linked to the real-time spindle speed data and input into the cutting torque calculation module. The real-time spindle cutting torque is calculated using the formula M = F × r, where F is the spindle axial cutting force component and r is the tool radius. Next, the cutting state turning point is determined based on the Z-axis cutting force component and cutting torque variation curve, and four cutting state identifiers are set, corresponding to light load, medium load, heavy load, and retraction stages. The spindle front and rear bearing temperatures (thermocouple temperature measurement accuracy does not exceed ±0.2°C) and vibration signals (acceleration range ±10g, sampling frequency 5000Hz) are synchronously detected. The spindle bearing state data is obtained using an over-limit judgment method. The spindle load factor is defined based on the bearing state and the spindle cutting torque. The calculation method is spindle load factor = cutting torque / bearing load limit. After obtaining the spindle load coefficient, the current spindle speed and the three-axis feed speed are extracted, and the matching degree is scored according to the matching relationship table between the speed and feed. The scoring range is 0 to 1. The higher the matching degree, the larger the feed rate can be set. Based on the matching degree, the response delay of the connection between the bed and the spindle is analyzed. The dynamic stiffness coefficient is calculated by measuring the ratio of the spindle speed change lag time to the feed speed change delay time, and it is set to float between 0.5 and 2.0. Then, the maximum feed speed upper limit of the servo motor of each axis is set according to the dynamic stiffness coefficient. The maximum feed speed of the X-axis, Y-axis, and Z-axis is limited to 15m / min, 12m / min, and 10m / min, respectively. Then, the rate of change of the torque and cutting force frequency distribution density in the tool force trajectory is analyzed, and the peak interval standard deviation is used to identify the degree of tool wear. The tool wear coefficient is calculated by combining the cumulative processing time and the load integral value, and input into the wear assessment unit. The resonance frequency characteristic values ​​and smoothed cutting force data obtained in step S1 are extracted using the Fourier transform method to extract the frequency domain distribution spectrum. The peak frequency points are compared with the natural frequency of the system to obtain the cutting vibration stability score, which ranges from 0 to 100. The tool wear coefficient and the cutting vibration stability score are jointly used to calculate the preliminary feed rate coefficient. The linear combination is based on the weight coefficients of 0.7 and 0.3, and the result is limited to between 0.4 and 1.2. This initial coefficient is multiplied by the basic feed rate of the X, Y, and Z axes (unit: mm / min) to obtain the three-axis feed coordination speed. The feed coordination coefficient is defined by the ratio between the maximum and minimum values. This coefficient is compared with the aforementioned feed rate upper limit to obtain the constrained feed rate coefficient, and its final value is limited to the range of 0.3 to 1.0.In order to meet the specific workpiece processing requirements, the constrained feed rate coefficient is corrected in combination with the preset surface roughness standard, and the quality optimization feed rate is reversed based on the relationship curve between the roughness target value Ra and the cutting speed. Then, the target efficiency is set according to the processing cycle requirements. Under the premise of meeting the Ra value, the quality optimization feed rate is adjusted and the efficiency balance factor is generated. The final optimized feed rate coefficient is output by multiplying the two.

[0044] Step S5: Generate a comprehensive control instruction based on the corrected position instruction, the thermal deformation compensation amount of each axis, and the optimized feed rate coefficient; construct a machining trajectory of the engraving and milling machine based on the optimized feed rate coefficient and the resonance frequency characteristic value, and perform optimized interpolation to obtain smoothed trajectory parameters; coordinately drive and control the servo motors of each axis based on the comprehensive control instruction and the smoothed trajectory parameters to obtain actual position feedback data of each axis;

[0045] In the embodiment of the present invention, the corrected position instruction output in step S2 and the thermal deformation compensation values ​​of the X-axis, Y-axis, and Z-axis obtained in step S3 are first numerically summed in axial order to generate a thermal compensation position instruction. Before being introduced into the servo control unit, the instruction needs to be combined with the temperature distribution of the guide rail area in the three-dimensional temperature field model, based on the guide rail material being high carbon steel and the linear expansion coefficient being 1.1×10 ―5 / ℃ parameter condition, the linear thermal expansion calculation formula ΔL=α·ΔT·L is used to simulate the thermal elongation of each linear guide pair (where the effective length of the guide is set to 700mm), and then the thermal elongation is superimposed on the thermal compensation position instruction to form the guide rail compensation position instruction. Subsequently, according to the optimized feed rate coefficient obtained in step S4, the guide rail compensation position instruction is speed modulated. The modulation method is to multiply the instruction value of each axis by the feed rate coefficient of the corresponding axis to form a speed modulation position instruction; this instruction is further combined with the workbench thermal deformation to perform coordinate calibration compensation. The thermal deformation is provided by the workbench thermal deformation simulation results. The origin of the machining coordinate system is corrected according to the X, Y, and Z directions in the three-dimensional coordinate system to obtain the thermal compensation value of the workpiece coordinate system. The speed modulation position instruction and the workpiece coordinate system thermal compensation value are superimposed and corrected in three-dimensional coordinates to generate the final comprehensive control instruction. During the machining trajectory generation process, a base trajectory velocity is set based on the optimized feedrate coefficient. Based on this velocity, trajectory points are derived from the tool center trajectory and the curvature is calculated. Continuous segments with curvature values ​​between 0.1 and 0.5 are extracted. The curvature difference is used to determine the severity of trajectory changes, and high-curvature segments are designated as key areas for interpolation optimization. Each trajectory segment is connected point by point using a piecewise interpolation method, and the control points of the starting and ending segments are smoothed using B-spline processing to ensure first- and second-order derivative continuity at each connection point. Combined with the resonant frequency eigenvalues ​​obtained in step S1, the trajectory velocity variation frequency is analyzed, and vibration frequency bands close to the system frequency response are excluded. If the difference between the frequency component of a trajectory segment and the resonant frequency is less than 10Hz, the segment is designated as a high-vibration risk segment. The vibration excitation frequency is reduced by adjusting the trajectory velocity gradient and interpolation density. Dynamic response coordination among multiple axes is also performed for this trajectory segment. Using the position changes of the three-axis motion trajectory at the same timestamp as a reference, inter-axis synchronization scheduling is performed. By comparing the differences in velocity and acceleration of each axis, the synchronization optimization adjustment amount is calculated and applied to the position command to obtain the synchronization optimization trajectory parameters. The B-spline smooth trajectory parameters are weightedly fused with the synchronization optimization trajectory parameters, with the fusion weight set to 0.65 for the B-spline trajectory parameters and 0.35 for the synchronization trajectory parameters. This generates the fused optimization trajectory parameters. The trajectory parameters are used to compare and analyze the geometric errors. Combined with the preset surface quality requirements and target machining efficiency, the feed rate and spindle speed coordination relationship is reversed, assuming the error does not exceed ±0.005 mm, to obtain the spindle feed coordination parameters. The spindle feed coordination parameters are combined with the ball screw parameters (lead 5 mm, transmission efficiency 0.9) and the trajectory is adaptively adjusted using the transmission relationship formula V = n × p / 60 (where V is the feed rate, n is the motor speed, and p is the lead), ultimately forming the smoothed trajectory parameters.The integrated control instructions and smoothed trajectory parameters are input into the multi-axis servo control driver, which controls the ball screw servo motor of each axis to achieve precise drive. At the same time, the spindle feed speed is modulated through coordinated parameters to achieve synchronous motion control between the spindle and each axis. Finally, the actual position data of each axis is fed back by the encoder.

[0046] Step S6: Evaluate the control accuracy of the actual position feedback data of each axis to obtain position error data; update the comprehensive control instructions based on the position error data, and perform dual-drive coordinated control on the engraving and milling machine.

[0047] In this embodiment of the present invention, the actual position coordinate data collected by the servo motors of each axis during the machining process in step S5 is first compared one-to-one with the smoothed trajectory parameters. The instantaneous position deviation values ​​of the X, Y, and Z axes are calculated using the difference calculation formulas ΔX = Xf - Xa, ΔY = Yf - Ya, and ΔZ = Zf - Za (where Xf, Yf, and Zf are the desired trajectory coordinates and Xa, Ya, and Za are the actual coordinates). All deviation data are recorded at a sampling interval of 0.5ms to form a position deviation data sequence. This sequence is then analyzed using standard deviation analysis and sliding mean processing to determine the stability of the deviation fluctuations. If the standard deviation of the three-axis deviation exceeds 0.003mm, a systematic error determination process is triggered. Based on this, the deviation data is statistically analyzed for positive and negative frequency by setting a fixed threshold (±0.005mm). If the deviation is biased toward one side with a frequency exceeding 60%, it is considered a systematic error; otherwise, it is considered a random error. After processing the two types of errors separately, the error characteristic parameter sets are calculated separately. The systematic error uses the periodic offset fitting method to determine the correction increment, and the random error uses the three-point weighted average method to extract the fluctuation center value as the dynamic fine-tuning amount. According to the above characteristic parameters, a closed-loop correction operation is performed on the comprehensive control instruction. The systematic correction increment is directly added to the trajectory position reference value, and the random dynamic fine-tuning amount is buffered and adjusted by the slope change of the speed instruction to generate a corrected comprehensive control instruction. Based on the correction instruction, the torque coordination calculation is performed on the main drive motor and preload motor configured at both ends of the ball screw of each axis. The total drive torque is set to no more than 90% of the rated torque, and the torque ratio of the main motor and the auxiliary motor is set based on the 30% sharing ratio of the preload motor to obtain the anti-backlash drive torque distribution data. This data is input into the servo controller as a dual-drive control command, which drives the main motor and preload motor to implement torque output, adjusting the loads on both ends in real time and synchronously, thereby generating a dual-drive synchronous control signal. This signal further controls the magnetic levitation micropositioner installed on each axis guide rail for submicron compensation. The magnetic levitation micropositioner uses the principle of electromagnetic suspension to achieve frictionless displacement control. The control signal is precisely modulated by a bidirectional PID current drive module with an adjustment resolution of 0.1μm and a response time of less than 2ms. This control method enables detailed correction of the end position coordinates. Ultimately, the compensated displacement signal output by the magnetic levitation micropositioner is superimposed with the original command displacement to form the final position control command, which is rewritten into the servo system's actuator for real-time position closed-loop control.

[0048] Through multi-source data fusion and intelligent modeling, this invention achieves comprehensive optimization of the servo position control system for milling machines, significantly improving their machining accuracy and dynamic response capabilities. First, by incorporating screw temperature, load, and multi-node temperature data from the entire machine, it effectively addresses the issues of insufficient compensation for gap errors and structural deformation caused by thermal expansion, load variations, and nonlinear thermal deformation, enhancing the system's robustness to thermal environment fluctuations. Second, by combining cutting force and resonance characteristic information, it enables dynamic adaptive adjustment of feed rate and trajectory smoothing interpolation, significantly improving trajectory smoothness and machining efficiency during milling and reducing trajectory deviations and surface defects caused by vibration. Furthermore, by performing real-time accuracy assessment and closed-loop error updates based on the actual position feedback of each axis, the system further enhances the response speed and accuracy of position control, ensuring stable operation under complex machining conditions. Finally, through a dual-drive coordinated control approach, the limitations of traditional control strategies in large dynamic loads and high-precision synchronous drive are effectively overcome, achieving higher-precision and more stable servo control performance, and overall improving the comprehensive performance and intelligence level of milling machines for high-speed, high-precision machining tasks.

[0049] Preferably, step S1 includes the following steps:

[0050] Step S11: performing temperature detection on the X-axis ball screw, the Y-axis ball screw, and the Z-axis ball screw of the engraving and milling machine to obtain real-time temperature data of the screw of each axis;

[0051] Step S12: detecting the axial load of each axis in the engraving and milling machine through a current sensor to obtain real-time load data of each axis;

[0052] Step S13: collecting the temperatures of the engraving and milling machine bed, column, spindle box and workbench through semiconductor temperature sensors to obtain multi-node temperature distribution data;

[0053] Step S14: detecting the cutting force of the engraving and milling machine by a three-axis force sensor to obtain cutting force data;

[0054] Step S15: performing spectrum analysis on the machine tool vibration of the engraving and milling machine through a vibration sensor to obtain a resonance frequency characteristic value.

[0055] In the embodiment of the present invention, at least two K-type thermocouple temperature sensors are installed on the outer wall of the ball screw nut seat of the X-axis, Y-axis, and Z-axis at equal intervals along the length of the screw. The sensor accuracy is controlled within the range of ±0.1°C, and the sampling period is set to 1 second. The temperature signal is transmitted to the data acquisition module via the RS485 bus, and the temperature changes of each axis screw during operation are recorded in real time, forming independent real-time temperature data for each axis. A Hall current sensor with a range of ±20A and an output sensitivity of 40mV / A is configured in series in the power supply circuit of each axial servo motor. The current fluctuation of each axis during the cutting process is monitored in milliseconds. The current value is then combined with the torque constant of the servo motor (unit: Nm / A) and the ball screw lead (unit: mm) and substituted into the axial load calculation formula: load = (current × motor torque constant) ÷ screw lead, and the real-time axial load data of the X, Y, and Z axes are calculated. A total of 32 thermistor-type semiconductor temperature sensors are arranged in typical heat-sensitive parts of the engraving and milling machine bed, column, spindle box and workbench. Each type of structure has no less than 8 measuring points. The measurement point layout is set according to the structural symmetry and heat conduction path. The sensor accuracy is controlled within ±0.2℃. I 2 The C bus communicates with the temperature acquisition module, sampling at least 60 times per minute to generate multi-node temperature distribution data for the four structural components of the bed, column, spindle box, and worktable. A three-dimensional temperature distribution matrix is ​​then constructed by combining the spatial coordinates of each measuring point. After completing temperature data acquisition, the system proceeds to step S14, where a three-dimensional piezoelectric force sensor is installed in the contact area between the spindle shank and the tool clamp. The force sensor should have a measuring range of 0 to 5000N and a sensitivity of at least 5pC / N. During cutting, the cutting force components in the X, Y, and Z directions are recorded with a time resolution of 0.5ms. The cutting force data is then synchronously acquired and cached using a signal amplifier connected to a high-speed data acquisition card. Three-axis MEMS acceleration vibration sensors are fixedly installed at the front end of the engraving and milling machine spindle box and the end of the three-axis guide rail. The frequency response range is required to be no less than 0.5Hz to 5000Hz. The signal is connected to the 24-bit AD converter through the analog channel. Vibration sampling is carried out when the machine tool is running at no load and in the speed variable frequency band. The sampling period is set to 1ms, and the sampling duration is no less than 10 seconds. The obtained vibration acceleration data is processed by fast Fourier transform to extract the frequency point with the maximum amplitude in the frequency domain, and its frequency value is recorded as the resonant frequency characteristic value of the structural system.

[0056] The present invention provides a solid data foundation for subsequent servo control optimization and error compensation by performing multi-dimensional and high-precision collection of temperature, load, cutting force and vibration data of key parts of the engraving and milling machine. Real-time monitoring of screw temperature and load helps to fully grasp the dynamic changes of screw thermal elongation and mechanical state, provide high-reliability input for predicting gap errors, and improve the accuracy of the compensation model; the acquisition of multi-node temperature field data of the whole machine enables visualization and modeling of structural thermal deformation, and enhances the timeliness and accuracy of thermal error modeling and compensation; the acquisition of cutting force information reflects the real-time interaction between the tool and the workpiece, which helps to achieve dynamic feed adjustment, improve processing efficiency and reduce surface defects; the vibration spectrum analysis results reveal the resonance characteristics of the machine tool structure, which provides a key basis for avoiding excitation frequency, suppressing trajectory oscillation and improving stability. Overall, this data acquisition strategy significantly enhances the environmental perception and dynamic response capabilities of the control system, and provides key support for achieving high-precision, high-stability and high-efficiency intelligent processing of engraving and milling machines.

[0057] Preferably, step S2 includes the following steps:

[0058] Step S21: filtering the real-time temperature data of the screw rod to obtain smoothed temperature data; normalizing the real-time load data to obtain standardized load data;

[0059] Step S22: Calculating the thermal expansion coefficient of the ball screw based on the smoothed temperature data and the normalized load data to obtain a thermal expansion influencing factor;

[0060] Step S23: detecting the transmission clearance reference value between the ball screw and the screw nut pair of each axis in the engraving and milling machine, and performing load deformation analysis on the screw nut pair to obtain the load deformation coefficient;

[0061] Step S24: constructing a neural network clearance prediction model based on the thermal expansion influence factor and the load deformation coefficient, and training the model to obtain a screw rod reverse clearance prediction model;

[0062] Step S25: Predicting the ball screw backlash of each axis in real time based on the ball screw backlash prediction model to obtain a dynamic backlash prediction value;

[0063] Step S26: Calculating the compensation amount of each axis based on the dynamic clearance prediction value to obtain a clearance compensation value;

[0064] Step S27: performing direction determination on the gap compensation value to obtain a forward compensation value and a reverse compensation value;

[0065] Step S28: Pre-compensation control is performed on the servo position of each axis before direction change based on the forward compensation value and the reverse compensation value to obtain a corrected position instruction.

[0066] In the embodiment of the present invention, the real-time temperature data of each axis screw collected in step S11 is first input into the filtering processing unit, and a weighted smoothing operation is performed on the five adjacent sampling points using the five-point sliding average method. The weight coefficients are set to 0.1, 0.2, 0.4, 0.2, and 0.1 to eliminate sudden interference and generate smoothed temperature data. Then, the real-time load data obtained in step S12 is normalized. The reference value is set to the rated axial load of 5000N of each axis screw under the maximum load state. The normalization adopts the standard linear transformation formula and limits the output value range to between 0 and 1 to obtain the standardized load data. According to the ΔT value in the smoothed temperature data and the linear expansion coefficient of the screw material GCr15 of 1.2×10 ―5 / °C, and an effective screw length of 800mm, the thermal expansion calculation formula ΔL = α·ΔT·L was used. Combined with standardized load data, a thermal expansion influence factor was constructed through linear superposition, with the temperature expansion component weighted as 0.6 and the load component weighted as 0.4, for coefficient fusion. A laser displacement sensor with a displacement measurement resolution of 0.001mm was used to measure the minimum starting displacement difference between each screw-nut pair in the no-load and reverse-drive states, recording it as the transmission clearance reference value. A static pressure loading device with a loading force of 1000N was used to apply axial loading to the nut pair. By comparing the displacement changes before and after loading, the load-deformation coefficient was calculated using the formula K = ΔL / F (where K is the deformation coefficient, ΔL is the displacement, and F is the loading force). A two-dimensional parameter mapping relationship diagram was constructed using equal weights for the thermal expansion influence factor and the load-deformation coefficient. Manual adjustment was used to correspond to historically measured actual clearance changes, forming a ball screw reverse clearance trend comparison table. This table was then used as the basis for interpolation to predict the real-time clearance during operation. The combined value of the current thermal expansion influencing factor and the load deformation coefficient is detected in real time, and the corresponding reverse clearance prediction value is found by comparing it with the trend comparison table, with the value range limited to 0.003mm to 0.015mm. The above dynamic clearance prediction value is directly used as the clearance compensation amount of the ball screw of each axis, and the dynamic clearance compensation value of the three axes is calculated separately and written into the compensation buffer area. By reading the current movement direction signal of the servo motor collected by the encoder, it is determined whether the screw is in the forward drive or reverse drive state. If it is forward drive, the forward compensation value is set to 0, and the reverse compensation value is equal to the clearance compensation value; if it is reverse drive, the forward compensation value is equal to the clearance compensation value, and the reverse compensation value is 0, realizing direction separation judgment. The forward compensation value and the reverse compensation value after the above direction judgment are superimposed on the basic position instruction, and vector addition processing is performed to generate a corrected position instruction.

[0067] The present invention achieves high-precision dynamic compensation for ball screw reverse clearance through multi-step intelligent processing and modeling, effectively improving the positioning accuracy and responsiveness of the milling machine in complex machining environments. Filtering and normalization enhance the stability and comparability of the original temperature and load data, reducing the interference of data fluctuations on model accuracy. The precise extraction of thermal expansion influencing factors and load deformation coefficients makes the compensation model more physically meaningful and predictive. The clearance prediction model constructed through a neural network has a strong fitting capability for nonlinear and multivariable relationships, and can respond in real time to the combined effects of screw thermal deformation and mechanical deformation, thereby more accurately predicting the trend of reverse clearance changes. Further direction differentiation and pre-compensation control of the compensation value enable the system to correct the position command in advance, avoiding error accumulation caused by feedback lag, and significantly improving the foresight and dynamic accuracy of servo control. Overall, this solution effectively solves the problems of reverse clearance compensation response lag, poor adaptability, and insufficient accuracy in traditional methods, providing key technical support for achieving high-speed, high-precision positioning and stable machining of milling machines.

[0068] Preferably, step S3 includes the following steps:

[0069] Step S31: constructing a three-dimensional digital twin temperature field based on multi-node temperature distribution data to obtain a three-dimensional temperature field model;

[0070] Step S32: performing a digital twin thermal deformation simulation of the guide rail mounting surface of the engraving and milling machine bed based on the three-dimensional temperature field model to obtain the thermal deformation of the bed;

[0071] Step S33: performing a thermal deformation simulation calculation of the connection between the column and the bed of the engraving and milling machine based on the three-dimensional temperature field model to obtain the thermal tilt deformation of the column;

[0072] Step S34: performing a thermal displacement simulation of the centerline position of the spindle of the engraving and milling machine based on the three-dimensional temperature field model to obtain the thermal displacement of the spindle;

[0073] Step S35: performing a thermal deformation simulation calculation of the workbench flatness based on the three-dimensional temperature field model to obtain the thermal deformation of the workbench;

[0074] Step S36: Calculate the X-axis thermal deformation compensation based on the bed thermal deformation and the spindle thermal displacement, calculate the Y-axis thermal deformation compensation based on the bed thermal deformation, the column thermal tilt deformation and the spindle thermal displacement, and calculate the Z-axis thermal deformation compensation based on the column thermal tilt deformation and the spindle thermal displacement, and finally form the thermal deformation compensation of each axis.

[0075] In the embodiment of the present invention, the temperature data collected by 12 semiconductor temperature sensors DS18B20 arranged on the engraving and milling machine body, the temperature data collected by 8 temperature sensors arranged on the column, the temperature data collected by 6 temperature sensors arranged on the spindle box, and the temperature data collected by 4 temperature sensors arranged on the workbench are spatially mapped to establish a three-dimensional rectangular coordinate system with the left front corner of the machine tool bed as the origin. The real-time temperature value of each temperature sensor is subjected to three-dimensional spatial interpolation calculation according to the X-coordinate, Y-coordinate, and Z-coordinate of the sensor installation position. The radial basis function interpolation method is used to expand the discrete 30 temperature measurement point data into a continuous temperature field distribution containing 10,000 grid nodes to form a three-dimensional digital twin temperature field model. Based on the temperature distribution data of the bed area in the three-dimensional temperature field model, a three-dimensional solid model of the bed guide rail mounting surface is established by ANSYS. The bed material is set to HT300 gray cast iron, the elastic modulus is set to 110 GPa, the Poisson's ratio is set to 0.26, and the linear thermal expansion coefficient is set to 1.1×10 ―5 / ℃, the temperature data of the bed area in the three-dimensional temperature field model is applied to the finite element model as the thermal load boundary condition, and the deformation displacement of the bed guide rail mounting surface under temperature change is calculated through thermal-structural coupling analysis. The thermal deformation displacement of the bed guide rail mounting surface along the X, Y, and Z directions is extracted as the thermal deformation of the bed; the temperature distribution data of the column area in the three-dimensional temperature field model is input into the column thermal deformation simulation model, and the column material parameters are set to ductile iron QT450-10, the elastic modulus is set to 169GPa, the Poisson's ratio is set to 0.275, and the linear thermal expansion coefficient is set to 1.2×10 ―5 / ℃, by constraining the degrees of freedom of the connection surface between the bottom surface of the column and the bed and releasing the rotational degrees of freedom of the top surface of the column, the thermal stress distribution and deformation state of the column under the action of the non-uniform temperature field are calculated, and the angular deflection of the top surface of the column relative to the bottom surface is extracted. The thermal tilt deformation of the column is calculated by multiplying the angular deflection by the column height of 800mm. The data of the six temperature measurement points in the spindle box area in the three-dimensional temperature field model are input into the spindle thermal displacement calculation program. The spindle material is set to 40Cr alloy steel, the elastic modulus is set to 206GPa, and the linear thermal expansion coefficient is set to 1.2×10 ―5 / ℃, by calculating the temperature difference between the front and rear bearing seats of the spindle, combined with the spindle length 320mm and the thermal expansion coefficient, the axial thermal elongation of the spindle is calculated, and the thermal displacement offset of the spindle centerline in the XY plane is calculated according to the uneven temperature distribution of the spindle box. The axial thermal elongation and radial thermal displacement offset are combined to form the thermal displacement of the spindle; the data of the four temperature measurement points in the worktable area in the three-dimensional temperature field model are fitted into the temperature distribution function of the worktable surface by the least squares method. The worktable material is set to HT300 gray cast iron. By calculating the thermal deformation difference caused by the temperature difference of the four corner points of the worktable, the bilinear interpolation method is used to calculate the thermal deformation of any point on the worktable surface, and the maximum value of the flatness deviation of the worktable surface is extracted as the thermal deformation of the worktable; by The X-axis thermal deformation compensation is calculated by vector synthesis of the directional component and the X-direction component of the spindle thermal displacement, which is ΔX=ΔX_bed+ΔX_spindle. The Y-direction component of the bed thermal deformation, the projection component of the column thermal tilt deformation in the Y direction, and the Y-direction component of the spindle thermal displacement are vector synthesized to calculate the Y-axis thermal deformation compensation as ΔY=ΔY_bed+ΔY_column+ΔY_spindle. The Z-direction component of the column thermal tilt deformation and the Z-direction component of the spindle thermal displacement are vector synthesized to calculate the Z-axis thermal deformation compensation as ΔZ=ΔZ_column+ΔZ_spindle. Finally, the thermal deformation compensation of the X-axis, Y-axis, and Z-axis of the engraving and milling machine are ΔX, ΔY, and ΔZ, respectively, and recorded as the thermal deformation compensation of each axis.

[0076] The present invention comprehensively improves the geometric accuracy control capability of the engraving and milling machine under thermal environment changes by constructing a refined three-dimensional digital twin temperature field and combining it with thermal deformation simulation of key structural parts. The three-dimensional temperature field model can truly reflect the thermal distribution characteristics of the machine tool under the influence of multi-point temperature rise, provide accurate boundary conditions for subsequent thermal deformation simulation, and significantly improve the credibility of thermal error modeling; virtual simulation of the thermal deformation of key components such as the bed, column, spindle and workbench can effectively identify and predict the structural displacement, tilt and warping trends during the processing process, and provide a quantitative basis for the compensation of each axis; by comprehensively calculating the composite effect of the thermal deformation of different components on the motion accuracy of each axis, the precise distribution of axial thermal compensation is achieved, and the trajectory deviation and part size error caused by thermal deformation are significantly reduced; this strategy enhances the geometric stability of the engraving and milling machine under continuous processing, heavy-load cutting or complex thermal field conditions, and provides an important guarantee for achieving high-precision and high-consistency processing. It is especially suitable for micro-structure manufacturing and precision mold processing scenarios where thermal-induced errors account for a significant proportion.

[0077] Preferably, step S4 includes the following steps:

[0078] Step S41: performing real-time filtering on the cutting force data to obtain smoothed cutting force data; calculating the spindle cutting torque based on the smoothed cutting force data to obtain the real-time cutting torque of the spindle; identifying the cutting state of the engraving and milling machine based on the smoothed cutting force data to obtain a cutting state identifier;

[0079] Step S42: detecting the temperature and vibration of the spindle bearing in the engraving and milling machine to obtain spindle bearing status data; evaluating the spindle load based on the real-time cutting torque of the spindle and the spindle bearing status data to generate a spindle load coefficient; analyzing the spindle speed and pre-feed speed matching based on the spindle load coefficient to obtain a speed feed matching degree;

[0080] Step S43: Analyzing the dynamic response of the engraving and milling machine bed and spindle based on the speed feed matching degree to obtain a dynamic stiffness coefficient; setting the upper limit of the feed speed of the servo motor of each axis in the engraving and milling machine based on the dynamic stiffness coefficient to obtain a feed speed constraint value;

[0081] Step S44: evaluating the degree of wear of the cutting edge of the tool in the engraving and milling machine based on the cutting state identifier to obtain a tool wear coefficient;

[0082] Step S45: performing cutting vibration frequency domain analysis based on the resonance frequency characteristic value and the smoothed cutting force data to obtain cutting vibration stability data;

[0083] Step S46: Calculating a feed rate reference value based on the tool wear coefficient and cutting vibration stability data to obtain an initial feed rate coefficient;

[0084] Step S47: Coordinate and optimize the feed speeds of the X-axis, Y-axis, and Z-axis based on the initial feed rate coefficient to obtain the feed coordination coefficient of each axis; perform feed rate optimization calculation based on the feed speed constraint value and the feed coordination coefficient of each axis to obtain the constrained feed rate coefficient;

[0085] Step S48: Optimize and adjust the surface quality of the constrained feed rate coefficient, calculate the efficiency balance, and generate the optimized feed rate coefficient.

[0086] In an embodiment of the present invention, the cutting force raw data collected by the three-axis force sensor Kistler9257B installed on the spindle box of the engraving and milling machine is subjected to Butterworth low-pass filtering, the filtering frequency is set to 500Hz, and the cut-off frequency is set to 100Hz to obtain smooth cutting force data, and the tangential component Ft in the smooth cutting force data is multiplied by the tool radius r to calculate the spindle real-time cutting torque T=Ft×r, when the cutting force resultant F_resultant is greater than the set threshold value 200N, it is marked as a heavy cutting state, when the cutting force resultant is less than 50N, it is marked as a light cutting state, and when the cutting force resultant is between 50N and 200N, it is marked as a normal cutting state, and the generated cutting state identifiers are 1, 2, and 3 corresponding to heavy cutting, normal cutting, and light cutting states respectively; the spindle bearing temperature and vibration acceleration are respectively collected by the temperature sensor PT100 and the acceleration sensor PCB352C33 installed on the front and rear bearing seats of the spindle. When the bearing temperature exceeds 65°C or the effective value of the vibration acceleration exceeds 5m / s 2The spindle bearing status data is marked as abnormal when the spindle bearing status data is abnormal. The spindle load coefficient η = T / 120 is calculated by dividing the real-time cutting torque T of the spindle by the rated torque of the spindle 120N·m. When the ratio n / vf of the spindle speed n to the feed speed vf is within the range of 800 to 1200rpm / (mm / min), the speed feed matching degree is set to 1.0. When it exceeds this range, it decreases linearly to 0.5 according to the degree of deviation. By measuring the displacement response of the engraving and milling machine bed under different cutting forces, the dynamic stiffness coefficient K_dynamic = ΔF / Δδ of the bed-spindle system is calculated, where ΔF is the change in cutting force and Δδ is the corresponding change in displacement. When the dynamic stiffness coefficient is less than 50N / μm, the servo motors of each axis are adjusted. The upper limit of the machine feed speed is set to 8000mm / min. When the dynamic stiffness coefficient is greater than 100N / μm, the upper limit of the feed speed is set to 15000mm / min. When it is between 50-100N / μm, the feed speed constraint value is calculated according to linear interpolation; the tool wear coefficient is calculated according to the cutting state identifier and the cumulative cutting time. When the cutting state identifier is 1, the wear coefficient growth rate is set to 0.02 / hour, when the identifier is 2, the growth rate is set to 0.01 / hour, and when the identifier is 3, the growth rate is set to 0.005 / hour. The tool wear coefficient K_wear is calculated cumulatively according to the corresponding growth rate starting from the initial value 0; the smoothed cutting force data and the resonance frequency characteristic value obtained in step S15 are input into In the fast Fourier transform processing program, the spectrum distribution of the cutting force signal is calculated. When the deviation between the cutting force frequency component and the machine tool resonance frequency is less than 5Hz, the cutting vibration stability data is marked as unstable. When the deviation is greater than 20Hz, it is marked as stable. When it is between 5 and 20Hz, it is marked as critical. The tool wear coefficient K_wear and the cutting vibration stability data are substituted into the formula F_base = 1.0-0.5×K_wear-0.3×V_instability to calculate the feed rate base value, where V_instability is the vibration instability coefficient. V_instability = 0 in stable state and V_instability = 0 in critical state. When ty = 0.5 and V_instability = 1.0 in the unstable state, the initial feed rate coefficient F_base is obtained. The feed coordination coefficient of each axis is obtained by multiplying the initial feed rate coefficient of the X-axis, Y-axis, and Z-axis by the inter-axis coordination factors of 0.9, 1.0, and 0.8, respectively. The feed coordination coefficient of each axis is compared with the corresponding feed speed constraint value. When the coordinated feed speed exceeds the constraint value, the feed rate coefficient is adjusted to the ratio of the constraint value divided by the base feed speed to obtain the constrained feed rate coefficient. The constrained feed rate coefficient is adjusted according to the preset workpiece surface roughness requirement Ra ≤ 1.6 μm. When the predicted surface roughness value exceeds 1.6 μm, the feed rate coefficient is multiplied by the correction factor 0.8. Get the quality optimized feed rate, calculate the efficiency balance factor by dividing the quality optimized feed rate by the preset processing efficiency requirement coefficient 1.2, and multiply the quality optimized feed rate by the efficiency balance factor to get the final optimized feed rate coefficient.

[0087] The present invention realizes adaptive optimization control of the feed speed of the engraving and milling machine by introducing multi-source dynamic characteristic data such as cutting force, spindle status, tool wear and vibration, effectively improving the stability, processing quality and overall efficiency of the processing process. The filtered cutting force data improves the accuracy of dynamic feature recognition, providing a stable basis for subsequent torque calculation and cutting state judgment; the analysis of spindle load and speed-feed matching enables the feed strategy to respond to the actual operating load of the equipment in real time, preventing overload and efficiency loss; the introduction of the dynamic stiffness coefficient enables the system to perceive the dynamic response capability of the structure, thereby avoiding processing errors or vibration excitation caused by insufficient stiffness; the integrated assessment of tool wear and cutting vibration stability enhances the comprehensive judgment of tool health and processing vibration risks, helping to prevent tool damage and suppress processing instability; the multi-level adjustment mechanism of the feed rate realizes the coordinated control of the speed of each axis and the adaptive constraint of the speed upper limit, making the processing process smoother and the trajectory smoother; ultimately, through the balance optimization between surface quality and efficiency, not only the surface integrity of the parts is improved, but also the production cycle requirements are taken into account, and the overall intelligent transformation of the engraving and milling machine feed control from passive response to active optimization is realized, which greatly enhances the processing adaptability and control flexibility under complex working conditions.

[0088] It is particularly important that step S48 includes the following steps:

[0089] Step S481: Optimizing the surface quality of the constrained feed rate based on the preset workpiece surface roughness requirement to obtain a quality-optimized feed rate;

[0090] Step S482: Calculating the efficiency balance based on the preset machining efficiency requirement and the quality optimized feed rate to obtain an efficiency balance factor;

[0091] Step S483: Determine the optimized feed rate coefficient based on the quality optimized feed rate and the efficiency balance factor.

[0092] In the embodiment of the present invention, the surface roughness requirement of the preset workpiece Ra≤1.6μm is used as the quality control benchmark, and the surface roughness prediction formula is used. Calculate the surface roughness prediction value under the current constraint feed rate coefficient, where f is the feed rate set to 0.1mm / r, r is the tool nose radius set to 0.8mm, vf is the feed rate, and n is the spindle speed. When the Ra_prediction value exceeds 1.6μm, the constraint feed rate coefficient is multiplied by the quality correction factor K_quality=1.6 / Ra_prediction for adjustment. When the Ra_prediction value is less than 1.0μm, the constraint feed rate coefficient is multiplied by the quality improvement factor 1.2 for optimization. When the Ra_prediction value is within the range of 1.0-1.6μm, the constraint is maintained. The feed rate coefficient remains unchanged, and the quality optimization feed rate F_quality is obtained through the above adjustment process; the preset processing efficiency requirement is set to the number of workpieces completed per hour N_target = 12 pieces, and the actual processing time T_actual = T_standard / F_quality is calculated based on the current quality optimization feed rate F_quality, where T_standard is the standard processing time set to 300 seconds / piece. The actual number of workpieces completed per hour is calculated based on the actual processing time N_actual = 3600 / T_ actual, the efficiency balance factor is calculated through the efficiency balance calculation formula η_balance = N_actual / N_target, when η_balance is greater than 1.2, the efficiency balance factor is limited to 1.2, when η_balance is less than 0.8, the efficiency balance factor is limited to 0.8, ensuring that the efficiency balance factor is in the range of 0.8 to 1.2; the quality optimization feed rate F_quality and the efficiency balance factor η_balance are weightedly fused and calculated, and the optimized feed rate coefficient is determined by the weighted average formula F_optimized = α×F_quality+β×η_balance×F_quality, where the quality weight coefficient α is set to 0.6 and the efficiency weight coefficient β is set to 0.4. When the calculated F_optimized exceeds the system maximum feed rate limit of 1.5, the optimized feed rate coefficient is limited to 1.5. When F_optimized is less than the system minimum feed rate limit of 0.3, the optimized feed rate coefficient is limited to 0.3, and finally the optimized feed rate coefficient in the range of 0.3 to 1.5 is obtained.

[0093] By introducing the dual constraints of workpiece surface quality and machining efficiency, the present invention achieves refined adjustment and multi-objective optimization of the feed rate, significantly improving the process adaptability and comprehensive performance of the engraving and milling machine under different machining tasks. Surface quality optimization adjustment can ensure that the feed rate suppresses the formation of cutting vibration marks and tool marks while meeting the preset roughness standard, thereby improving the consistency and accuracy of the machined surface. The introduction of the efficiency balance factor achieves a dynamic balance between the contradictory relationship between machining time and surface quality, avoiding a significant decrease in efficiency due to excessive pursuit of quality, or sacrificing surface accuracy due to increased efficiency. The optimized feed rate coefficient finally generated comprehensively considers the two core goals of quality and efficiency, allowing the machining parameters to automatically adapt and intelligently adjust under different process requirements, thereby significantly enhancing the system's practicality and flexible control capabilities in high-precision, high-efficiency manufacturing scenarios.

[0094] Preferably, in step S5, the comprehensive control instructions of the servo position controllers of each axis in the engraving and milling machine are output based on the corrected position instructions, the thermal deformation compensation amount of each axis and the optimized feed rate coefficient, including:

[0095] Perform data fusion processing based on the corrected position instruction and the thermal deformation compensation amount of each axis to obtain the thermal compensation position instruction;

[0096] Based on the three-dimensional temperature field model, the thermal deformation simulation analysis of the linear guide pair is carried out to obtain the thermal elongation of the guide rail;

[0097] Based on the thermal compensation position command and the thermal elongation of the guide rail, the linear guide pair is compensated for its accuracy to obtain the guide rail compensation position command;

[0098] Based on the guide rail compensation position command and the optimized feed rate coefficient, the servo position controller speed is modulated to obtain the speed modulation position command;

[0099] Calculate the thermal displacement compensation of the origin of the workpiece coordinate system based on the thermal deformation of the worktable to obtain the thermal compensation value of the workpiece coordinate system;

[0100] The workpiece coordinates are corrected based on the speed modulation position command and the thermal compensation value of the workpiece coordinate system, and multi-axis synchronous timing control is performed to obtain a comprehensive control command.

[0101] In the embodiment of the present invention, the corrected position command (X_corrected, Y_corrected, Z_corrected) obtained in step S2 is numerically added to the thermal deformation compensation amount (ΔX, ΔY, ΔZ) of each axis obtained in step S3 to calculate the thermal compensation position command as X_thermal = X_corrected + ΔX, Y_thermal = Y_corrected + ΔY, and Z_thermal = Z_corrected + ΔZ. Then, the three-dimensional temperature field data in step S3 is input into the linear guide pair thermal deformation calculation program. The guide rail material is set to 45# steel, and the linear thermal expansion coefficient is 1.2×10 ―5 / ℃, the thermal expansion of the X-axis guide rail with a length of 1200mm under temperature change ΔT_X is calculated as L_X = 1200×1.2×10 ―5 ×ΔT_X, thermal expansion of the Y-axis guide rail with a length of 800 mm L_Y=800×1.2×10 ―5 ×ΔT_Y, the thermal expansion of the Z-axis guide rail is 600mm L_Z=600×1.2×10 ―5×ΔT_Z, the thermal compensation position instruction and the guide rail thermal elongation are superimposed to obtain the guide rail compensation position instruction X_guide=X_thermal+L_X, Y_guide=Y_thermal+L_Y, Z_guide=Z_thermal+L_Z, and then the guide rail compensation position instruction is multiplied by the optimized feed rate coefficient F_optimized obtained in step S4, and the modulation speed of each axis is calculated by the speed modulation formula V_modulated=V_original×F_optimized to obtain the speed modulation position instruction including the position component and the speed component. At the same time, the thermal displacement compensation of the origin of the workpiece coordinate system is calculated according to the thermal deformation of the worktable ΔZ_table in step S3, and the origin of the workpiece coordinate system is adjusted from (0,0,0) to (0,0,ΔZ_table) through the coordinate transformation matrix to obtain the thermal compensation value of the workpiece coordinate system as (0,0,ΔZ_table). Then the speed modulation position instruction and the workpiece coordinate system thermal compensation value are subjected to coordinate system transformation operation, and the matrix operation [X_final Y_finalZ_final]=[X_modulatedY_modulated Z_modulated]+[0 0ΔZ_table] to correct the workpiece coordinates. Finally, the multi-axis synchronous timing control program synchronizes the corrected position instructions of each axis according to the interpolation period of 1ms to ensure that the position instructions of the X-axis, Y-axis, and Z-axis are consistent in time. The comprehensive control instructions are generated by timing synchronization calculation X_sync(t)=X_final×sin(2πt / T), Y_sync(t)=Y_final×sin(2πt / T), and Z_sync(t)=Z_final×sin(2πt / T), where T is the interpolation period set to 1ms and t is the current time step. The final output is a comprehensive control instruction containing position, velocity, and acceleration information.

[0102] The present invention achieves high-precision dynamic generation of servo control instructions for each axis of the engraving and milling machine through multi-source data fusion and layered thermal compensation strategy, significantly improving the control accuracy and processing consistency of the system in complex thermal-mechanical environments. The generation of thermal compensation position instructions effectively integrates the position correction and the influence of thermal deformation of each axis, making the control instructions closer to the actual state of the machine tool and fundamentally reducing the position drift caused by thermal errors. The simulation analysis and compensation processing of the thermal deformation of the linear guide pair further improves the positioning accuracy of the machine tool at the end of the motion chain and solves the geometric error problem caused by the thermal elongation of the guide rail. The speed modulation mechanism is combined with the optimized feed rate coefficient to enable the servo control system to adaptively adjust the response speed according to the actual load and process status, improving the dynamic performance and trajectory smoothness of the processing. At the same time, the thermal displacement of the workpiece coordinate system origin is corrected to ensure high consistency between the processing characteristics and the design coordinate system, enhancing the geometric matching of the finished product size. Finally, through multi-axis synchronous control, timing coordination is achieved, avoiding the coupling amplification of thermal errors under multi-axis linkage, and overall optimizing the execution accuracy, system stability and processing quality of the engraving and milling machine during high-speed and high-precision processing.

[0103] Preferably, in step S5, constructing a machining trajectory of the engraving and milling machine based on the optimized feed rate coefficient and the resonance frequency characteristic value, and performing optimized interpolation includes:

[0104] The original machining trajectory speed of the engraving and milling machine is calibrated based on the optimized feed rate coefficient, and the geometric characteristics of the machining trajectory of the engraving and milling machine are analyzed based on the original machining trajectory speed to obtain the trajectory curvature characteristic data;

[0105] Construct and train a neural network trajectory optimization model based on trajectory curvature feature data to obtain a trajectory optimization neural network model;

[0106] Based on the trajectory optimization neural network model, intelligent optimization interpolation of the machining trajectory is performed to obtain preliminary optimized trajectory parameters;

[0107] Based on the preliminary optimized trajectory parameters and resonance frequency eigenvalues, vibration avoidance interpolation processing is performed, and the B-spline curve smoothing processing of the trajectory transition point is performed to obtain the B-spline smooth trajectory parameters;

[0108] The frequency response of the engraving and milling machine bed and spindle is analyzed based on the resonance frequency eigenvalue pair. The vibration risk of each trajectory segment of the engraving and milling machine is evaluated based on the frequency response analysis results to obtain the vibration risk level.

[0109] Based on the vibration risk level, the dynamic response analysis of the linear guide pair of each axis of the engraving and milling machine is carried out. Based on the dynamic response analysis results, the multi-axis coordinated optimization of the motion trajectory of each axis of the engraving and milling machine is carried out to obtain the multi-axis coordinated trajectory parameters.

[0110] Based on the multi-axis coordinated trajectory parameters, the motion synchronization of the servo motors of each axis in the engraving and milling machine is optimized to obtain the synchronous optimized trajectory parameters;

[0111] The geometric accuracy and motion continuity are verified through B-spline smooth trajectory parameters and synchronous optimization trajectory parameters, the processing efficiency and quality are balanced, the spindle feed speed is coordinated and the ball screw transmission characteristics are optimized to generate smooth trajectory parameters.

[0112] In the embodiment of the present invention, the optimized feed rate coefficient generated by step S4 is first limited to between 0.4 and 1.2, and the corrected processing speed of each trajectory segment is calibrated by multiplying the coefficient with the unit feed rate of the original G code trajectory point point by point. Then, the coordinate derivative operation of the continuous trajectory points is performed by the central difference method, and the curvature radius between adjacent points is calculated, and then the curvature feature data of the processing path in three-dimensional space is extracted, and the analysis interval is limited to every 5 trajectory points. Then, the curvature change amplitude of each segment is counted using a sliding window, the curvature mutation point is identified, and the 10 trajectory points before and after it are extracted as the high curvature area, and then The trajectory of the high curvature area is reconstructed by interpolation smoothing. The cubic B-spline curve is used to refit the trajectory between these points to ensure that the number of control points of each B-spline is not less than 4 and the node spacing does not exceed 0.2mm. Combined with the resonance frequency characteristic value obtained by FFT analysis of the vibration sensor, the velocity spectrum of the preliminary reconstructed trajectory is compared with the resonance frequency. All trajectory segments with velocity excitation components near the resonance frequency band (within the frequency difference range of ±5%) are interpolated and reconstructed. The specific method is to add a velocity buffer segment in the segment and adjust the acceleration curve to avoid the excitation frequency. B-spline continuous second-order derivative processing is applied to the points to ensure that there is no speed mutation at the corner connection and the smoothness of the trajectory; then, based on the frequency response function of the bed-spindle system, the structural dynamics method is used to input equal-amplitude sinusoidal feed excitation to different trajectory segments and extract the output response. After quantifying the response amplitude, it is divided into three vibration risk levels: low, medium, and high, and dynamic response adjustment strategies are set respectively. Among them, the high-risk segment needs to insert a speed reduction zone and the interpolation nodes are encrypted to 0.1mm; on this basis, the dynamic response analysis of the linear guide pair of each axis is carried out. The analysis method is based on the simplified finite element guide structure model to input the velocity and acceleration data of each segment of the trajectory. After evaluating its stiffness-mass coupling response, the feed speed timing of each axis is fine-tuned to form the multi-axis coordinated trajectory parameters; then the multi-axis coordinate instructions are synchronized and optimized, and a time series phase alignment method is used to ensure that the synchronization deviation of each axis reaching the interpolation point does not exceed 1ms, generating the synchronous optimized trajectory parameters; finally, the B-spline smooth trajectory parameters are fused with the synchronous optimized trajectory parameters, and a weighted average strategy is used to integrate the velocity, acceleration and position data, and the geometric error and trajectory continuity indicators are calculated respectively. The geometric error shall not exceed 0.02mm, and the maximum jump of the first-order derivative of the velocity curve continuity shall not exceed 10mm / s.2 After the conditions are met, speed matching correction is performed based on the spindle rated speed range (3000-18000 rpm) and the maximum allowable feed speed of the ball screw of each axis (no more than 40 m / min), and trajectory speed correction is performed based on the screw transmission stiffness, and finally smooth trajectory parameters for servo position control are generated.

[0113] This invention integrates optimized feed rate with machine tool resonance characteristics to construct a high-precision, vibration-resistant, and continuously smooth machining trajectory, significantly improving the trajectory control capability and machining quality of milling machines under high-speed and complex paths. By combining trajectory curvature characteristics with a neural network model, the trajectory interpolation process is intelligently optimized, automatically adjusting motion instructions based on path geometry changes to avoid error accumulation and impact loads caused by sudden trajectory changes. The introduction of resonance frequency characteristics enables vibration avoidance and smoothing of trajectory transition points, effectively suppressing excitation responses during trajectory execution and improving trajectory continuity and stability. Trajectory segment vibration risk assessment combined with guideway dynamic response analysis enables the system to optimize trajectory structure and adjust motion strategies in advance during potential resonance segments, reducing vibration interference risks at the source. Multi-axis trajectory coordination and synchronous optimization control ensure motion consistency and response alignment during the linkage of each axis, improving spatial path restoration accuracy. The resulting smooth trajectory parameters not only achieve an optimal balance between geometric accuracy and dynamic stability, but also dynamically coordinate the spindle feed rate based on path complexity and improve the load transmission characteristics of the ball screw, significantly enhancing the trajectory execution quality, surface finish consistency, and overall machining efficiency of the milling machine.

[0114] Of particular importance is the verification of geometric accuracy and motion continuity through B-spline smoothing and synchronous optimization of trajectory parameters, the balance of machining efficiency and quality, the coordination of spindle feed speed, and the optimization of ball screw transmission characteristics, including:

[0115] Based on the B-spline smooth trajectory parameters and the synchronous optimization trajectory parameters, trajectory fusion processing is performed to obtain the fusion optimized trajectory parameters;

[0116] Based on the fusion optimization trajectory parameters, the trajectory geometric accuracy and motion continuity are verified and analyzed to obtain trajectory quality evaluation data;

[0117] Based on the preset processing efficiency requirements and trajectory quality evaluation data, the precision and efficiency balance is optimized to obtain the balanced optimized trajectory parameters;

[0118] Based on the balanced optimization trajectory parameters, the spindle speed and the feed speed of each axis are coordinated and matched to obtain the spindle feed coordination parameters;

[0119] The ball screw transmission characteristics in the engraving and milling machine are optimized based on the spindle feed coordination parameters to generate smooth trajectory parameters.

[0120] In the embodiment of the present invention, the B-spline smooth trajectory parameters obtained in step S5 and the synchronous optimization trajectory parameters are first matched with each other in a time step manner. At each time point, the three sets of data of corresponding position, velocity, and acceleration are fused using the weighted average method, and the weight coefficients are set to 0.6, 0.3, and 0.1, respectively, to generate the fused optimized trajectory parameters. Then, based on the coordinate continuity and velocity change rate of each segment of the trajectory point in the fused trajectory, the geometric accuracy and motion continuity analysis are performed. The specific operation is to import the fused trajectory in the CAD / CAM environment, calculate the maximum deviation value within each 100mm trajectory length segment by segment and limit it to not more than ±0.02mm, and at the same time calculate the first-order velocity derivative between adjacent trajectory points and limit its change rate to not more than 15mm / s. 2 , thereby obtaining the trajectory quality evaluation data; then set the processing efficiency target as a feed path length of not less than 500mm per minute, and adjust it in combination with the trajectory segment length and speed stability index in the evaluation data, perform linear buffer interpolation processing on the segment with drastic speed changes, and perform B-spline local reconstruction processing on the segment with large geometric deviation, and finally form a balanced optimized trajectory parameter; based on the average feed speed data of each segment of this trajectory, and combined with the spindle rated output power range of 2.2kW to 5.5kW and the tool processing radius, the spindle speed range is inversely solved according to the cutting linear speed formula v=π·D·n / 1000, and the feed speed of each segment is matched with the spindle speed. Insert a speed adjustment transition section into the section required by the process to generate the spindle feed coordination parameters; input the spindle feed coordination parameters and the acceleration and load data of the trajectory section corresponding to the balance optimization trajectory parameters of each axis into the controller, use the guide rail friction compensation coefficient (limited to 0.015) and the ball screw transmission stiffness (limited to 100-120N / μm) for calculation, perform transmission matching analysis on each trajectory, and correct the speed overshoot and position tracking error caused by the dynamic response lag of the ball screw by adjusting the acceleration rise slope and feed rate fine-tuning control to ensure that the transmission response delay of each axis during the trajectory execution does not exceed 2ms, and finally generate smooth trajectory parameters.

[0121] The present invention achieves a comprehensive improvement in the geometric accuracy, motion continuity and execution coordination of the engraving and milling machine processing trajectory by integrating trajectory geometry optimization and synchronous motion control strategy. Trajectory fusion processing improves the smoothness of the overall path and the local dynamic matching capability, effectively reducing the risk of impact and speed jitter caused by trajectory switching and transfer points; through geometric accuracy and continuity verification, it ensures that the path can still accurately restore the design contour under high dynamic response conditions, improving the contour fidelity and surface quality of the final product; the introduction of an accuracy and efficiency balance mechanism optimizes the machine tool beat while ensuring high-quality trajectory execution, significantly improving the system operation efficiency; the coordinated matching of the spindle speed and the multi-axis feed speed further enhances the synchronous response performance of the motion chain, keeps the spindle cutting stability consistent with the feed rhythm, and avoids processing defects caused by speed incoordination; at the same time, the adaptive optimization of the ball screw transmission characteristics based on the trajectory characteristics effectively suppresses the screw transmission nonlinear error caused by trajectory fluctuations, achieves higher positioning stability and load response consistency, and overall improves the trajectory execution accuracy, processing consistency and system robustness of the engraving and milling machine under high-speed and high-precision conditions.

[0122] Preferably, in step S5, the coordinated drive control of the servo motors of each axis in the engraving and milling machine based on the comprehensive control instructions and the smoothing trajectory parameters includes:

[0123] Based on the comprehensive control instructions, the ball screw servo motor of each axis in the engraving and milling machine is position controlled and driven to obtain the motor angle instruction data of each axis;

[0124] Based on the smoothed trajectory parameters, the spindle motor speed in the engraving and milling machine is coordinated and controlled to obtain the spindle speed synchronization instruction;

[0125] Based on the motor angle command data of each axis and the spindle speed synchronization command, the multi-axis linkage control of the engraving and milling machine bed guide rail is carried out to obtain the synchronization data of the motion trajectory of each axis;

[0126] Collect the position feedback signal of the high-precision encoder installed on the ball screw of each axis of the engraving and milling machine to obtain the pulse feedback data of the encoder of each axis;

[0127] Based on the synchronous data of each axis motion trajectory and the pulse feedback data of each axis encoder, the actual position of the engraving and milling machine worktable is calculated to obtain the actual position coordinate data of each axis;

[0128] The actual position coordinate data of each axis is monitored and analyzed in real time to obtain the actual position feedback data of each axis.

[0129] In the embodiment of the present invention, first, according to the comprehensive control instruction generated in step S5, the target position instructions of the three axes of X, Y and Z are extracted, and the target displacement of each axis in the smoothed trajectory parameters and the corresponding timestamp information are combined. The trajectory is discretized to a sampling period of 1ms by interpolation, and the expected position, speed and acceleration in each sampling period are calculated based on the third-order position planning curve. The motion controller outputs the result to the servo driver, drives the ball screw servo motors of the three axes in the engraving and milling machine, and generates accurate motor angle instruction data. At the same time, the spindle motor control According to the trajectory speed corresponding to the cutting segment and the blank segment in the smoothing trajectory parameters, the linear speed conversion formula v = π·D·n / 1000 (where the tool diameter D is limited to Φ6mm, and v is taken from the average speed of the trajectory segment) is used to reversely infer the synchronous speed that the spindle should reach. The slope control is performed based on the limit that the spindle acceleration time constant τ does not exceed 0.3s to form the spindle speed synchronization instruction. Subsequently, the rotation angle instruction data of the three-axis motor and the spindle synchronous speed signal are input into the motion control bus together to construct a multi-axis coordinated control scheme with the time base as the synchronization reference. The absolute timestamp comparison method is used for linkage stepping processing to generate synchronous data of the motion trajectory of each axis including position, speed, and acceleration; then, in each sampling period, the high-precision grating encoder (accuracy not less than 0.1μm) installed at the end of the three-axis ball screw outputs a pulse feedback signal, and the encoder pulse number is converted into the actual rotation angle through differential processing, and then multiplied by the lead of the screw (X axis 5mm / r, Y axis 5mm / r, Z axis 4mm / r) to convert the linear displacement to obtain the actual displacement of each axis; then the actual displacement data is synchronously read The Cartesian coordinate solution formula is taken and inserted, and the coordinate conversion is performed in combination with the machine origin offset value and the screw installation direction definition to obtain the actual position coordinate data of each axis at the current time point; finally, the position coordinate data is compared with the trajectory synchronization data cycle by cycle to calculate the speed difference, position deviation and acceleration change rate. The servo controller monitors the motion state in real time. If any of the three indicators exceeds the set threshold (the position error threshold is ±0.01mm, the speed difference threshold is ±5mm / s, and the acceleration change rate threshold is ±50mm / s), the position error is detected and the position error is detected. 2 ), an adjustment signal is immediately issued and recorded as actual position feedback data.

[0130] The present invention achieves high-precision synchronous control of the servo motors of each axis of the engraving and milling machine through the coordinated drive of comprehensive control instructions and smoothing trajectory parameters, effectively improving the coordination of machine tool movement and processing accuracy. The ball screw motor angle instruction based on position control drive ensures the precise positioning of each axis and improves the execution accuracy of the processing path; the coordinated control of the spindle motor speed achieves synchronous matching of the spindle and feed motion, ensuring the stability and processing efficiency of the cutting process; the multi-axis linkage control achieves smooth movement of complex spatial trajectories through the linkage management of the bed guide rails, reducing mechanical vibration and dynamic impact; high-precision encoder feedback captures the motion status of each axis in real time, ensuring high responsiveness and accuracy of position data; the coordinate solution combining motion trajectory synchronization data and encoder feedback improves the real-time monitoring accuracy of the actual position; the overall precise position control and dynamic status monitoring of the engraving and milling machine worktable are achieved, enhancing the stability of the processing process, repeatable positioning capability and system fault warning capability, and significantly improving the operating reliability and processing quality of the engraving and milling machine in a high-speed and high-precision processing environment.

[0131] Preferably, step S6 includes the following steps:

[0132] Step S61: performing real-time error calculation on the position accuracy of the worktable of the engraving and milling machine based on the actual position feedback data of each axis to obtain the position deviation data of each axis;

[0133] Step S62: classifying the position deviation data of each axis into systematic errors and random errors to obtain a set of error characteristic parameters;

[0134] Step S63: adaptively correcting and updating the integrated control instruction based on the error characteristic parameter set to obtain a corrected integrated control instruction;

[0135] Step S64: performing torque distribution calculation on the main drive motor and the preload motor at both ends of the ball screw of each axis in the engraving and milling machine based on the modified comprehensive control instruction to obtain anti-backlash drive torque distribution data;

[0136] Step S65: using the anti-backlash drive torque distribution data to perform dual-drive coordinated control on the main drive motor and the preload motor of each axis in the engraving and milling machine to obtain a dual-drive synchronous control signal;

[0137] Step S66: Based on the dual-drive synchronous control signal, sub-micron precision compensation control is performed on the magnetic suspension micro-positioner integrated on the guide rails of each axis in the engraving and milling machine to obtain the final position control instruction.

[0138] In the embodiment of the present invention, first, the actual position coordinate data of each axis obtained by position solution in step S5 is subjected to point-by-point subtraction processing with the target position data in the smoothed trajectory parameters, and sampling is performed once every 1ms. The real-time position deviation data of the X-axis, Y-axis, and Z-axis in the whole process are calculated using the difference formula, and the out-of-tolerance segment is screened with an accuracy threshold of 0.01mm; then all the deviation data are subjected to discrete Fourier transform to extract the frequency domain components within a trajectory cycle (set to every 1000ms), and the deviation amount with a frequency lower than 1Hz and periodicity is classified as a systematic error, and the deviation amount with a frequency higher than 5Hz and irregular amplitude change is classified as a random error, and the maximum value, mean value, and standard deviation of the two types of errors are calculated respectively to form an error characteristic parameter set; then, according to the axial direction corresponding to the maximum value of the systematic error in the error characteristic parameter set, the command position value of the axis in the original integrated control instruction is linearly offset corrected, and the correction value is equal to the negative value of the mean value of the systematic error. At the same time, for the axial direction with obvious random error, the high acceleration section of the trajectory is interpolated. The speed buffer points are inserted, and the length of each buffer point is limited to no more than 2mm, thus forming a corrected comprehensive control instruction. Subsequently, based on the acceleration and friction load parameters of each axis motion segment in the corrected comprehensive control instruction, combined with the ball screw lead, rolling resistance and backlash value, the output torque distribution ratio of the main drive motor and the preload motor is calculated. The output ratio of the main drive motor is set to 75%, and the output ratio of the preload motor is set to 25%. The output result is the anti-backlash drive torque distribution data. This data is used to transmit it to the servo controller of each axis respectively to achieve dual-drive synchronous control, where the maximum response time difference between the two motors does not exceed 1ms, and the output current synchronization error does not exceed 3%, forming a dual-drive synchronous control signal. Finally, according to the control signal, the magnetic suspension micropositioner integrated in the guide pair of each axis is driven in real time (drive resolution is 0.1μm, response time is less than 0.5ms). Its control method is to follow the corrected position drift curve with equal amplitude to perform reverse displacement compensation, so that the actual position deviation of each axis converges to within the range of ±0.5μm, and the final position control instruction is generated.

[0139] The present invention realizes dynamic monitoring and precise identification of the position accuracy of the engraving and milling machine worktable through real-time error calculation and error feature classification, effectively distinguishes systematic errors from random errors, and improves the accuracy of error diagnosis; adaptive control instruction correction based on error features improves the response speed and adjustment accuracy of the servo system to position deviations, and enhances the intelligence level of the overall control; torque distribution calculation rationally distributes the load of the main drive motor and the preload motor, realizes efficient coordination of the anti-backlash drive, and significantly reduces the influence of mechanical gap on positioning accuracy; dual-drive coordinated control ensures the synchronous execution of the main drive and the preload drive, and improves the stability and response consistency of the driving force transmission; combined with the sub-micron precision compensation control of the magnetic levitation micropositioner, ultra-high precision adjustment of the guide rail movement is realized, which greatly reduces tiny vibrations and displacement errors, and ultimately improves the positioning stability and repeatability of the engraving and milling machine in high-precision processing, and significantly improves the processing quality and the reliability of equipment operation.

[0140] Preferably, the present invention further provides a servo position control system for an engraving and milling machine, which is used to execute the above-mentioned servo position control method for an engraving and milling machine. The servo position control system for an engraving and milling machine comprises:

[0141] Multi-source data acquisition module, used to obtain the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collect the multi-node temperature distribution data of the engraving and milling machine; obtain the cutting force data and resonance frequency characteristic value of the engraving and milling machine;

[0142] The intelligent backlash prediction and compensation module is used to build a screw backlash prediction model based on the screw's real-time temperature and load data, and generate backlash compensation values. Based on the backlash compensation values, pre-compensation control is performed before the servo position and direction of each axis are changed to obtain a corrected position command.

[0143] The digital twin thermal deformation simulation module is used to simulate the thermal deformation of the engraving and milling machine based on multi-node temperature distribution data and generate thermal deformation compensation for each axis;

[0144] Adaptive feed adjustment module, used to adjust the feed speed of the engraving and milling machine based on cutting force data to obtain the optimized feed rate coefficient;

[0145] The multi-axis coordinated control module is used to generate comprehensive control instructions based on the corrected position instructions, thermal deformation compensation of each axis, and optimized feed rate coefficients; construct the machining trajectory of the engraving and milling machine based on the optimized feed rate coefficients and resonance frequency characteristic values, and perform optimized interpolation to obtain smoothed trajectory parameters; based on the comprehensive control instructions and smoothed trajectory parameters, the servo motors of each axis are coordinated and controlled to obtain actual position feedback data of each axis;

[0146] The precision assessment feedback optimization module is used to evaluate the control precision of the actual position feedback data of each axis to obtain the position error data; based on the position error data, it updates the comprehensive control instructions and performs dual-drive coordinated control of the engraving and milling machine.

[0147] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is not limited by the above description. Therefore, it is intended that all changes that fall within the meaning and scope of the equivalent elements of the application documents are included in the present invention.

[0148] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A servo position control method for a milling machine, characterized in that: The following steps are involved: Step S1: Acquire the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; Collect multi-node temperature distribution data of the engraving and milling machine; obtain cutting force data and resonance frequency characteristic values ​​in the engraving and milling machine; Step S2: constructing a screw rod backlash prediction model based on the screw rod real-time temperature data and real-time load data, and generating a backlash compensation value; performing pre-compensation control before the servo position direction of each axis is changed based on the backlash compensation value to obtain a corrected position command; Step S3: performing thermal deformation simulation on the engraving and milling machine based on the multi-node temperature distribution data to generate thermal deformation compensation for each axis; Step S4: adjusting the feed speed of the engraving and milling machine based on the cutting force data to obtain an optimized feed rate coefficient; Step S5: Generate a comprehensive control instruction based on the corrected position instruction, the thermal deformation compensation amount of each axis, and the optimized feed rate coefficient; construct a machining trajectory of the engraving and milling machine based on the optimized feed rate coefficient and the resonance frequency characteristic value, and perform optimized interpolation to obtain smoothed trajectory parameters; coordinately drive and control the servo motors of each axis based on the comprehensive control instruction and the smoothed trajectory parameters to obtain actual position feedback data of each axis; Step S6: Evaluate the control accuracy of the actual position feedback data of each axis to obtain position error data; update the comprehensive control instructions based on the position error data, and perform dual-drive coordinated control on the engraving and milling machine.

2. The servo position control method for a milling machine according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: performing temperature detection on the X-axis ball screw, the Y-axis ball screw, and the Z-axis ball screw of the engraving and milling machine to obtain real-time temperature data of the screw of each axis; Step S12: detecting the axial load of each axis in the engraving and milling machine through a current sensor to obtain real-time load data of each axis; Step S13: collecting the temperatures of the engraving and milling machine bed, column, spindle box and workbench through semiconductor temperature sensors to obtain multi-node temperature distribution data; Step S14: detecting the cutting force of the engraving and milling machine by a three-axis force sensor to obtain cutting force data; Step S15: performing spectrum analysis on the machine tool vibration of the engraving and milling machine through a vibration sensor to obtain a resonance frequency characteristic value.

3. The servo position control method for a milling machine according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: filtering the real-time temperature data of the screw rod to obtain smoothed temperature data; normalizing the real-time load data to obtain standardized load data; Step S22: Calculating the thermal expansion coefficient of the ball screw based on the smoothed temperature data and the normalized load data to obtain a thermal expansion influencing factor; Step S23: detecting the transmission clearance reference value between the ball screw and the screw nut pair of each axis in the engraving and milling machine, and performing load deformation analysis on the screw nut pair to obtain the load deformation coefficient; Step S24: constructing a neural network clearance prediction model based on the thermal expansion influencing factor and the load deformation coefficient, and training the model to obtain a screw rod reverse clearance prediction model; Step S25: Predicting the ball screw backlash of each axis in real time based on the ball screw backlash prediction model to obtain a dynamic backlash prediction value; Step S26: Calculating the compensation amount of each axis based on the dynamic clearance prediction value to obtain a clearance compensation value; Step S27: performing direction determination on the gap compensation value to obtain a forward compensation value and a reverse compensation value; Step S28: Pre-compensation control is performed on the servo position of each axis before direction change based on the forward compensation value and the reverse compensation value to obtain a corrected position instruction.

4. The servo position control method for a milling machine according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: constructing a three-dimensional digital twin temperature field based on multi-node temperature distribution data to obtain a three-dimensional temperature field model; Step S32: performing a digital twin thermal deformation simulation of the guide rail mounting surface of the engraving and milling machine bed based on the three-dimensional temperature field model to obtain the thermal deformation of the bed; Step S33: performing a thermal deformation simulation calculation of the connection between the column and the bed of the engraving and milling machine based on the three-dimensional temperature field model to obtain the thermal tilt deformation of the column; Step S34: performing a thermal displacement simulation of the centerline position of the spindle of the engraving and milling machine based on the three-dimensional temperature field model to obtain the thermal displacement of the spindle; Step S35: performing a thermal deformation simulation calculation of the workbench flatness based on the three-dimensional temperature field model to obtain the thermal deformation of the workbench; Step S36: Calculate the X-axis thermal deformation compensation based on the bed thermal deformation and the spindle thermal displacement, calculate the Y-axis thermal deformation compensation based on the bed thermal deformation, the column thermal tilt deformation and the spindle thermal displacement, and calculate the Z-axis thermal deformation compensation based on the column thermal tilt deformation and the spindle thermal displacement, and finally form the thermal deformation compensation of each axis.

5. The servo position control method for a milling machine according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing real-time filtering on the cutting force data to obtain smoothed cutting force data; calculating the spindle cutting torque based on the smoothed cutting force data to obtain the real-time cutting torque of the spindle; identifying the cutting state of the engraving and milling machine based on the smoothed cutting force data to obtain a cutting state identifier; Step S42: detecting the temperature and vibration of the spindle bearing in the engraving and milling machine to obtain spindle bearing status data; evaluating the spindle load based on the real-time cutting torque of the spindle and the spindle bearing status data to generate a spindle load coefficient; analyzing the spindle speed and pre-feed speed matching based on the spindle load coefficient to obtain a speed feed matching degree; Step S43: Analyzing the dynamic response of the engraving and milling machine bed and spindle based on the speed feed matching degree to obtain a dynamic stiffness coefficient; setting the upper limit of the feed speed of the servo motor of each axis in the engraving and milling machine based on the dynamic stiffness coefficient to obtain a feed speed constraint value; Step S44: evaluating the degree of wear of the cutting edge of the tool in the engraving and milling machine based on the cutting state identifier to obtain a tool wear coefficient; Step S45: performing cutting vibration frequency domain analysis based on the resonance frequency characteristic value and the smoothed cutting force data to obtain cutting vibration stability data; Step S46: Calculating a feed rate reference value based on the tool wear coefficient and cutting vibration stability data to obtain an initial feed rate coefficient; Step S47: Coordinate and optimize the feed speeds of the X-axis, Y-axis, and Z-axis based on the initial feed rate coefficient to obtain the feed coordination coefficient of each axis; perform feed rate optimization calculation based on the feed speed constraint value and the feed coordination coefficient of each axis to obtain the constrained feed rate coefficient; Step S48: Optimize and adjust the surface quality of the constrained feed rate coefficient, calculate the efficiency balance, and generate the optimized feed rate coefficient.

6. The servo position control method for a milling machine according to claim 1, characterized in that: In step S5, the comprehensive control instruction is generated based on the corrected position instruction, the thermal deformation compensation amount of each axis and the optimized feed rate coefficient, including: Perform data fusion processing based on the corrected position instruction and the thermal deformation compensation amount of each axis to obtain the thermal compensation position instruction; Based on the three-dimensional temperature field model, the thermal deformation simulation analysis of the linear guide pair is carried out to obtain the thermal elongation of the guide rail; Based on the thermal compensation position command and the thermal elongation of the guide rail, the linear guide pair is compensated for its accuracy to obtain the guide rail compensation position command; Based on the guide rail compensation position command and the optimized feed rate coefficient, the servo position controller speed is modulated to obtain the speed modulation position command; Calculate the thermal displacement compensation of the origin of the workpiece coordinate system based on the thermal deformation of the worktable to obtain the thermal compensation value of the workpiece coordinate system; The workpiece coordinates are corrected based on the speed modulation position command and the thermal compensation value of the workpiece coordinate system, and multi-axis synchronous timing control is performed to obtain a comprehensive control command.

7. The servo position control method for a milling machine according to claim 1, characterized in that: In step S5, the machining trajectory of the engraving and milling machine is constructed based on the optimized feed rate coefficient and the resonance frequency characteristic value, and the optimized interpolation is performed, including: The original machining trajectory speed of the engraving and milling machine is calibrated based on the optimized feed rate coefficient, and the geometric characteristics of the machining trajectory of the engraving and milling machine are analyzed based on the original machining trajectory speed to obtain the trajectory curvature characteristic data; Construct and train a neural network trajectory optimization model based on trajectory curvature feature data to obtain a trajectory optimization neural network model; Based on the trajectory optimization neural network model, intelligent optimization interpolation of the machining trajectory is performed to obtain preliminary optimized trajectory parameters; Based on the preliminary optimized trajectory parameters and resonance frequency eigenvalues, vibration avoidance interpolation processing is performed, and the B-spline curve smoothing processing of the trajectory transition point is performed to obtain the B-spline smooth trajectory parameters; The frequency response of the engraving and milling machine bed and spindle is analyzed based on the resonance frequency eigenvalue pair. The vibration risk of each trajectory segment of the engraving and milling machine is evaluated based on the frequency response analysis results to obtain the vibration risk level. Based on the vibration risk level, the dynamic response analysis of the linear guide pair of each axis of the engraving and milling machine is carried out. Based on the dynamic response analysis results, the multi-axis coordinated optimization of the motion trajectory of each axis of the engraving and milling machine is carried out to obtain the multi-axis coordinated trajectory parameters. Based on the multi-axis coordinated trajectory parameters, the motion synchronization of the servo motors of each axis in the engraving and milling machine is optimized to obtain the synchronous optimized trajectory parameters; The geometric accuracy and motion continuity are verified through B-spline smooth trajectory parameters and synchronous optimization trajectory parameters, the processing efficiency and quality are balanced, the spindle feed speed is coordinated and the ball screw transmission characteristics are optimized to generate smooth trajectory parameters.

8. The servo position control method for a milling machine according to claim 1, characterized in that: In step S5, the coordinated drive control of the servo motors of each axis based on the comprehensive control instructions and the smoothed trajectory parameters includes: Based on the comprehensive control instructions, the ball screw servo motor of each axis in the engraving and milling machine is position controlled and driven to obtain the motor angle instruction data of each axis; Based on the smoothed trajectory parameters, the spindle motor speed in the engraving and milling machine is coordinated and controlled to obtain the spindle speed synchronization instruction; Based on the motor angle command data of each axis and the spindle speed synchronization command, the multi-axis linkage control of the engraving and milling machine bed guide rail is carried out to obtain the synchronization data of the motion trajectory of each axis; Collect the position feedback signal of the high-precision encoder installed on the ball screw of each axis of the engraving and milling machine to obtain the pulse feedback data of the encoder of each axis; Based on the synchronous data of each axis motion trajectory and the pulse feedback data of each axis encoder, the actual position of the engraving and milling machine worktable is calculated to obtain the actual position coordinate data of each axis; The actual position coordinate data of each axis is monitored and analyzed in real time to obtain the actual position feedback data of each axis.

9. The servo position control method for a milling machine according to claim 1, characterized in that: Step S6 includes the following steps: Step S61: performing real-time error calculation on the position accuracy of the worktable of the engraving and milling machine based on the actual position feedback data of each axis to obtain the position deviation data of each axis; Step S62: classifying the position deviation data of each axis into systematic errors and random errors to obtain a set of error characteristic parameters; Step S63: adaptively correcting and updating the integrated control instruction based on the error characteristic parameter set to obtain a corrected integrated control instruction; Step S64: performing torque distribution calculation on the main drive motor and the preload motor at both ends of the ball screw of each axis in the engraving and milling machine based on the modified comprehensive control instruction to obtain anti-backlash drive torque distribution data; Step S65: using the anti-backlash drive torque distribution data to perform dual-drive coordinated control on the main drive motor and the preload motor of each axis in the engraving and milling machine to obtain a dual-drive synchronous control signal; Step S66: Based on the dual-drive synchronous control signal, sub-micron precision compensation control is performed on the magnetic suspension micro-positioner integrated on the guide rails of each axis in the engraving and milling machine to obtain the final position control instruction.

10. A servo position control system for an engraving and milling machine, characterized in that: Used to execute the servo position control method for an engraving and milling machine according to claim 1, the servo position control system for an engraving and milling machine comprises: Multi-source data acquisition module, used to obtain the real-time temperature data and real-time load data of the screw of each axis in the engraving and milling machine; collect the multi-node temperature distribution data of the engraving and milling machine; obtain the cutting force data and resonance frequency characteristic value of the engraving and milling machine; The intelligent backlash prediction and compensation module is used to build a screw backlash prediction model based on the screw's real-time temperature and load data, and generate backlash compensation values. Based on the backlash compensation values, pre-compensation control is performed before the servo position and direction of each axis are changed to obtain a corrected position command. The digital twin thermal deformation simulation module is used to simulate the thermal deformation of the engraving and milling machine based on multi-node temperature distribution data and generate thermal deformation compensation for each axis; Adaptive feed adjustment module, used to adjust the feed speed of the engraving and milling machine based on cutting force data to obtain the optimized feed rate coefficient; The multi-axis coordinated control module is used to generate comprehensive control instructions based on the corrected position instructions, thermal deformation compensation of each axis, and optimized feed rate coefficients; construct the machining trajectory of the engraving and milling machine based on the optimized feed rate coefficients and resonance frequency characteristic values, and perform optimized interpolation to obtain smoothed trajectory parameters; based on the comprehensive control instructions and smoothed trajectory parameters, the servo motors of each axis are coordinated and controlled to obtain actual position feedback data of each axis; The precision assessment feedback optimization module is used to evaluate the control precision of the actual position feedback data of each axis to obtain the position error data; based on the position error data, it updates the comprehensive control instructions and performs dual-drive coordinated control of the engraving and milling machine.

Images