A method and system for optimizing the multi-axis linkage trajectory of a five-axis linkage machining center

By constructing a virtual machine tool model synchronized with the physical machine tool, the motion lag deviation of the five-axis machine tool is pre-calculated and compensated, solving the problem of delay in traditional feedback compensation, realizing high-precision five-axis linkage machining, and improving the trajectory accuracy and surface quality of high-speed machining.

CN122794929BActive Publication Date: 2026-11-10JONAK CNC EQUIPMENT (JIANGSU) CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202611247700.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-18
Publication Date
2026-11-10
Estimated Expiration
2046-08-18

AI Technical Summary

Technical Problem

When existing five-axis linkage machining centers perform high-speed machining, the dynamic response lag and compensation delay of the transmission chain lead to a decrease in trajectory accuracy, resulting in vibration marks and surface quality defects.

Method used

By constructing a virtual machine tool model that runs synchronously with the physical machine tool, the motion lag deviation caused by the dynamic characteristics of the mechanical transmission chain is pre-calculated, and a pre-compensation correction vector is generated. The motion commands are adjusted in real time to eliminate lag, thereby achieving feedforward compensation.

Benefits of technology

It significantly improves the trajectory accuracy and surface quality of five-axis machine tools during high-speed machining, effectively suppresses vibration marks and surface defects, and improves machining efficiency and machining quality of complex curved parts.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122794929B_ABST
    Figure CN122794929B_ABST
Patent Text Reader

Abstract

This application relates to the field of CNC machining technology and discloses a method and system for optimizing the multi-axis linkage trajectory of a five-axis linkage machining center. The method includes: acquiring the multi-axis motion command sequence of a five-axis machine tool; calling a preset virtual machine tool model to calculate the expected motion state of the execution end of the five-axis machine tool; comparing the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response lag deviation of each motion axis; generating a corresponding pre-compensation correction vector based on the physical response lag deviation; correcting the multi-axis motion command sequence based on the pre-compensation correction vector; and outputting the corrected multi-axis motion command sequence to the drive system of the five-axis machine tool. This ensures that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion command sequence is synchronized with the theoretical motion state, achieving feedforward compensation for dynamic errors, eliminating the delay problem of traditional feedback compensation, and improving the trajectory accuracy and surface quality of high-speed linkage machining.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of CNC machining technology, and more specifically, to a method and system for optimizing the multi-axis linkage trajectory of a five-axis linkage machining center. Background Technology

[0002] When performing high-speed, high-precision machining of complex curved surface parts, the final machining quality of a five-axis machining center highly depends on the trajectory accuracy of the multi-axis linkage. Traditional control methods typically perform one-time compensation during the trajectory planning stage based on the static error model calibrated at the machine tool's factory, or rely on position feedback from the servo system for real-time correction during machining. However, for large five-axis machine tools with complex structures and long transmission chains, this approach has significant limitations. The mechanical transmission chain of the machine tool, including components such as lead screws, gears, rotary tables, and oscillating heads, does not have static physical characteristics. During high-speed motion and frequent acceleration and deceleration, factors such as inertia, elasticity, friction, and backlash can cause dynamic, motion-state-related response lags and deformations, which cannot be accurately described by static models. Furthermore, compensation methods relying on real-time feedback suffer from time delays in signal acquisition, calculation, and command execution, resulting in compensation actions always lagging behind the actual occurrence of the error. This lag is particularly fatal in high-speed cutting and sections where the trajectory direction changes rapidly. It triggers so-called catch-up corrections, which not only fail to effectively eliminate errors but may also cause speed fluctuations and vibrations, ultimately leaving defects such as chatter marks and uneven light bands on the workpiece surface, severely affecting contour accuracy and surface quality. The root cause is that existing technology lacks a mechanism to anticipate and synchronously respond to the dynamic physical response of the transmission chain, resulting in an irreparable misalignment between theoretical commands and machine tool physical execution.

[0003] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides a multi-axis linkage trajectory optimization control method and system for a five-axis linkage machining center, which can solve the technical problem in the prior art where the trajectory accuracy of a five-axis machine tool decreases during high-speed machining due to the lag in the dynamic response and compensation delay of the transmission chain.

[0005] In a first aspect, this application provides a method for optimizing the multi-axis linkage trajectory control of a five-axis linkage machining center, including: Obtain the multi-axis motion command sequence of a five-axis machine tool; The preset virtual machine tool model is invoked, and the expected motion state of the execution end of the five-axis machine tool is calculated based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical characteristics of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. The expected motion state is compared with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. Generate a corresponding pre-compensation correction vector based on the physical response hysteresis deviation; The multi-axis motion command sequence is corrected based on the pre-compensation correction vector, and the corrected multi-axis motion command sequence is output to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion command sequence is synchronized with the theoretical motion state.

[0006] This technical solution utilizes a virtual machine tool model that runs synchronously with the physical machine tool to pre-calculate the motion lag and deviation caused by the dynamic characteristics of the mechanical transmission chain, and corrects the motion commands in advance accordingly. This achieves feedforward compensation for dynamic errors, fundamentally eliminating the delay problem of traditional feedback compensation. This allows the actual physical response of the machine tool to be highly synchronized with the theoretical commands, significantly improving the trajectory accuracy and surface quality of high-speed linkage machining.

[0007] Furthermore, the dynamic parameters characterizing the physical properties of the mechanical transmission chain include inertia parameters, stiffness parameters, and damping parameters; the steps of calling a preset virtual machine tool model and calculating the expected motion state of the execution end of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical properties of the mechanical transmission chain include: The multi-axis motion command sequence is input into the preset virtual machine bed model. Based on the multi-axis motion command sequence, inertial parameters, stiffness parameters and damping parameters, dynamic iterative calculation is performed to obtain the expected position and expected attitude of the execution end, which are taken as the expected motion state.

[0008] By introducing key dynamic parameters such as inertia, stiffness, and damping for iterative calculations, the model can more realistically simulate the force and deformation behavior of the machine tool transmission chain during motion, thereby improving the accuracy of predicting the expected motion state of the execution end and providing a reliable basis for accurately calculating hysteresis deviation.

[0009] Furthermore, after outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method also includes: Real-time acquisition of output torque data from the drive system; Extract the equivalent load component related to the machining load based on the output torque data; The stiffness parameters are dynamically weakened and compensated based on the equivalent load components to correct the elastic deformation of the five-axis machine tool under load.

[0010] By monitoring the motor torque in real time to infer the machining load and dynamically adjusting the stiffness parameters in the virtual machine tool model, the model can account for the structural elastic deformation caused by the cutting force, further improving the accuracy of the model prediction. Especially under heavy-load cutting conditions, it can more accurately compensate for the trajectory deviation caused by the load.

[0011] Furthermore, the steps for dynamically weakening and compensating for stiffness parameters based on equivalent load components include: Based on the equivalent load component, the matching stiffness attenuation coefficient is retrieved from the preset stiffness load mapping table, which records the stiffness attenuation coefficients corresponding to different loads. Multiplying the stiffness parameter by the stiffness attenuation coefficient yields the equivalent stiffness parameter of the five-axis machine tool under the current machining load, which is then used as the compensated stiffness parameter.

[0012] This technical solution provides an efficient engineering approach for dynamic stiffness compensation. By employing a pre-defined mapping table, the complex load-stiffness relationship is simplified to table lookup and multiplication operations, significantly reducing the complexity of real-time calculations and ensuring that the compensation algorithm can be completed quickly within the servo control cycle, thus guaranteeing the real-time performance of the control system.

[0013] Furthermore, before outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method also includes: Obtain the physical performance base information of the drive system, including the maximum torque limit of each motion axis motor, the maximum current limit of the driver, and the safe load-bearing capacity of the transmission components. Determine whether the corrected multi-axis motion command sequence reaches the physical output limit based on the physical performance base information.

[0014] This technical solution introduces a safety verification step for the machine tool's physical capabilities. Before outputting compensation commands, it is pre-determined whether they will exceed the limits of hardware such as motors and drives, avoiding commands that could damage the machine tool due to over-compensation, and ensuring the safety and robustness of the optimized control method.

[0015] Furthermore, when it is determined that the modified multi-axis motion command sequence has reached the physical output limit, the step of outputting the modified multi-axis motion command sequence to the drive system includes: Calculate the output gap ratio; Based on the output gap ratio, the feed commands of all motion axes at the same moment are reduced proportionally to obtain a reduced multi-axis motion command sequence. This ensures the geometric consistency of the multi-axis linkage trajectory without exceeding the physical output limit. The reduced multi-axis motion command sequence is then output to the drive system.

[0016] By proportionally reducing all linkage axes, rather than simply shaving and limiting the amplitude, the spatial motion trajectory shape of the tool center point is ensured to remain unchanged even when the speed is reduced to adapt to hardware limitations, thus maintaining the machining contour accuracy to the maximum extent while ensuring safety.

[0017] Furthermore, after outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method also includes: Real-time acquisition of the actual position feedback of the drive system; The actual position feedback is compared with the expected motion state to determine the position deviation characteristics; When the position deviation characteristics meet the preset trend increase condition, it is determined that the physical model parameters of the five-axis machine tool have drifted, and the model parameters of the preset virtual machine tool model are adjusted based on the position deviation characteristics so that they are refitted to the current physical model parameters of the five-axis machine tool.

[0018] By comparing model predictions with actual machine tool feedback, the slow changes in the physical characteristics of the machine tool caused by factors such as wear and temperature rise can be monitored, and the model parameters can be automatically adjusted accordingly. This ensures that the virtual machine tool model can always accurately reflect the real state of the physical machine tool during long-term operation, and maintains the lasting effectiveness of the compensation effect.

[0019] Furthermore, the steps for adjusting the model parameters of the preset virtual machine bed model based on the positional deviation characteristics include: The changes in the physical model parameters of the five-axis machine tool are calculated in reverse based on the characteristics of the positional deviation. The damping parameter or discrete time step integral parameter in the preset virtual machine bed model is adjusted in real time according to the changes in the physical model parameters.

[0020] By reverse calculation, macroscopic positional deviations are attributed to changes in microscopic physical parameters, and factors that significantly affect dynamic response, such as damping or integral parameters, are adjusted accordingly, making model calibration more accurate and efficient.

[0021] Furthermore, the steps of comparing the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis include: Calculate the physical response hysteresis deviations for the linear axis and the rotation axis respectively; For a linear axis, calculate the position lag deviation vector value between the expected position of the expected motion state and the theoretical position of the theoretical motion; For the rotation axis, calculate the attitude angle hysteresis vector value between the expected attitude angle of the expected motion state and the theoretical attitude angle of the theoretical motion.

[0022] This technical solution differentiates the calculation of deviations for different types of motion axes. Recognizing the differences in the physical descriptions of linear and rotational motion, position vectors and attitude angle vectors are used to quantify the deviations, making the deviation calculations more consistent with kinematic principles and resulting in more targeted and accurate compensations.

[0023] Secondly, this application also discloses a multi-axis linkage trajectory optimization control system for a five-axis linkage machining center, used to execute the aforementioned multi-axis linkage trajectory optimization control method for a five-axis linkage machining center, the system comprising: The instruction acquisition module is used to acquire the multi-axis motion instruction sequence of a five-axis machine tool; The state prediction module is used to call a preset virtual machine tool model and calculate the expected motion state of the execution end of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical characteristics of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. The deviation determination module is used to compare the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. The vector generation module is used to generate a corresponding pre-compensation correction vector based on the physical response hysteresis deviation. The instruction output module is used to correct the multi-axis motion instruction sequence based on the pre-compensation correction vector, and output the corrected multi-axis motion instruction sequence to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion instruction sequence is synchronized with the theoretical motion state.

[0024] This application provides a method and system for optimizing the multi-axis linkage trajectory of a five-axis linkage machining center. Its core advantage lies in constructing a virtual machine tool model with parameters identical to the physical machine tool model and synchronized with its operating cycle, thereby achieving accurate prediction and feedforward compensation of the machine tool's dynamic response. Compared to existing technologies, this application no longer passively waits for errors to occur before performing hysteresis correction. Instead, it proactively and in advance calculates the response hysteresis deviation that inevitably arises under high-speed motion due to the physical characteristics of the transmission chain, such as inertia, stiffness, and damping, and pre-corrects the original motion commands accordingly. This pre-compensation mechanism fundamentally solves the time delay problem of traditional feedback control, enabling the corrected commands to guide the actual motion state of the physical machine tool to be highly synchronized with the theoretical trajectory. Therefore, this application can significantly improve the contour accuracy of five-axis machine tools under high-speed and high-acceleration / deceleration conditions, effectively suppress vibration marks and surface defects, and ensure the machining quality of complex curved surface parts while improving machining efficiency. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a multi-axis linkage trajectory optimization control method for a five-axis linkage machining center, as provided in an embodiment of this application.

[0026] Figure 2 This is a schematic diagram of the structure of a multi-axis linkage trajectory optimization control system for a five-axis linkage machining center, provided in an embodiment of this application.

[0027] Labeling Explanation: 210, Instruction Acquisition Module; 220, State Prediction Module; 230, Deviation Determination Module; 240, Vector Generation Module; 250, Instruction Output Module. Detailed Implementation

[0028] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0029] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0030] In the aerospace manufacturing field, machining critical components with thin walls and complex curved surfaces, such as integral bladed disks, places extremely stringent requirements on machining accuracy and surface quality. The blade sections of integral bladed disks exhibit severe torsion, requiring high-speed, high-acceleration complex linkages between the rotary and linear axes of a five-axis machine tool during machining. In traditional machining methods, the CNC system directly drives each servo motor based on the theoretical toolpath generated by CAM software—a multi-axis motion command sequence. However, because the machine tool itself is a large and complex mechanical system, a long transmission chain exists between the motor issuing commands and the actual execution at the tool tip, including couplings, ball screws, gears and racks, bearings, guideways, and structures such as rotary tables and tilting heads. These components are not ideal rigid bodies; they exhibit inertia, elasticity, and friction. When the machine tool moves at high speeds, especially during rapid turns or acceleration / deceleration, inertia leads to sluggish motion, elasticity causes minute deformations and vibrations, and changes in friction introduce unstable resistance. These physical effects combine to cause the actual trajectory of the tool tip to always lag behind or deviate from the theoretically commanded trajectory. Although these deviations are minute, under high-speed cutting, they will leave fine vibration marks in the transition area of ​​the blade surface, or form uneven light bands on surfaces requiring high surface finish. These are direct manifestations of excessive contour errors. Existing compensation schemes that rely on encoder position feedback are always half a beat late due to the inherent time delay in signal transmission and processing, and cannot fundamentally eliminate this dynamic error.

[0031] Regarding this, firstly, see... Figure 1 This application provides a method for optimizing the multi-axis linkage trajectory control of a five-axis linkage machining center, the method comprising: S1. Obtain the multi-axis motion command sequence of the five-axis machine tool; S2. Call the preset virtual machine tool model, and calculate the expected motion state of the execution end of the five-axis machine tool according to the multi-axis motion command sequence and the dynamic parameters characterizing the physical characteristics of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. S3. Compare the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. S4. Generate the corresponding pre-compensation correction vector based on the physical response hysteresis deviation; S5. Based on the pre-compensation correction vector, the multi-axis motion command sequence is corrected, and the corrected multi-axis motion command sequence is output to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion command sequence is synchronized with the theoretical motion state.

[0032] The pre-defined virtual machine tool model can be understood as a digital twin running in a computer, completely corresponding to the physical machine tool. This model is not a simple three-dimensional geometric model, but a mathematical model encompassing the complete dynamic characteristics of the physical machine tool. It abstracts each motion axis of the machine tool (e.g., X, Y, Z linear axes and A, C rotary axes) into a set of physical equations describing its motion behavior. The model's parameters, such as the mass of each moving part, moment of inertia, transmission chain stiffness, and damping coefficients of bearings and guideways, are obtained through measurement and calibration of the actual physical machine tool, ensuring that the model's parameters are consistent with the physical model parameters of the physical machine tool. Furthermore, the operation of this virtual machine tool model is synchronized with the servo control system of the physical machine tool. That is, while the physical machine tool's servo controller performs a position loop control every extremely short interval (e.g., 1 millisecond), the virtual machine tool model also performs a calculation every 1 millisecond. This synchronization is the foundation for achieving accurate predictions.

[0033] Physical response hysteresis refers to the difference between the theoretically expected position and orientation of the tool at a given moment and the actual position and orientation predicted by the virtual machine tool model after considering various physical effects. This deviation is not a simple numerical value, but a vector containing deviation components of each motion axis. It quantitatively describes the degree of motion distortion caused by the physical limitations of the mechanical system.

[0034] The pre-compensation correction vector is a reverse vector generated based on the calculated physical response hysteresis deviation. Before the command is issued, the original command is corrected, informing the machine tool in advance how much additional movement is needed to exactly offset the impending physical response hysteresis deviation.

[0035] The overall process of the multi-axis linkage trajectory optimization control method for the five-axis linkage machining center is described in detail below.

[0036] First, the control system obtains the sequence of multi-axis motion commands to be executed from the upstream computer-aided manufacturing software or CNC program. This sequence is essentially a list of timestamps and corresponding five-axis coordinates. For example, at time point t1, the tool tip should be located at coordinates (x1, y1, z1, a1, c1); at time point t2, it should be located at (x2, y2, z2, a2, c2), and so on. This sequence represents the perfect trajectory that the tool should follow under ideal conditions, i.e., the theoretical motion state.

[0037] Next, before sending these instructions to the physical machine tool's drive system, the control system first sends the same instruction sequence to a pre-defined virtual machine tool model. Upon receiving the theoretical target position at each moment, the virtual machine tool model doesn't simply accept the result; instead, it performs an iterative calculation based on its internally established dynamic equations. This calculation simulates real-world phenomena: when an instruction requires a massed object to start moving from rest, the object needs to overcome inertia to accelerate; when an instruction requires it to turn at high speed, the transmission chain will undergo slight elastic deformation due to the force; and friction will also hinder the movement. The virtual machine tool model numerically solves these dynamic equations to calculate, under the current instruction and considering all these physical effects, the actual position and orientation that the machine tool's end effector (i.e., the tool tip) can reach at the end of the next servo cycle. This calculation result is the expected motion state.

[0038] As a specific example, the construction of a virtual machine tool model can be based on multibody dynamics theory. For instance, for a typical cradle-type five-axis machine tool, its three linear axes (X, Y, and Z), as well as the C-axis driving the table rotation and the A-axis driving the spindle oscillation, can be modeled as a series of rigid and flexible bodies connected by joints. The inertial parameters of each component, such as mass and moment of inertia tensors, can be accurately extracted from the CAD model of the machine tool design. The stiffness parameters of the transmission chain, such as the tensile and torsional stiffness of the ball screw, the meshing stiffness of the gear pair, and the support stiffness of the bearings, can be determined through finite element analysis or experimental measurement (e.g., applying known forces to the components and measuring their minute displacements with a laser interferometer). Damping parameters represent the energy dissipation in the system, such as friction and material resistance, and are usually identified by analyzing the attenuation characteristics of the vibration response after frequency sweep excitation of the machine tool. Substituting these parameters into the whole-machine dynamics equations based on Newton-Euler's equations or Lagrange's equations constitutes the core of the virtual machine tool model. When a motion command within a time step is input, the model solver (such as the fourth-order Runge-Kutta method) uses these equations to calculate the acceleration, velocity, and displacement of each axis, and finally obtains the expected motion state of the execution end.

[0039] After obtaining the expected motion state, the control system immediately compares it with the theoretical motion state at the same moment. This comparison process is performed axis by axis to determine the physical response hysteresis deviation of each motion axis. For example, at time t, if the theoretical command requires the X-axis position to be 100.000 mm, but the virtual machine model predicts that the actual X-axis position can only reach 99.998 mm, then the physical response hysteresis deviation of the X-axis is -0.002 mm. Similarly, if the theoretical command requires the C-axis to rotate to 90.000 degrees, but the model predicts that it can only rotate to 89.995 degrees, then the deviation of the C-axis is -0.005 degrees.

[0040] Subsequently, the system generates a corresponding pre-compensation correction vector based on the physical response hysteresis deviation of all axes. In the most direct implementation, this correction vector is the negative value of the hysteresis deviation. Continuing the example above, for a hysteresis of -0.002 mm on the X-axis, the generated correction component is +0.002 mm; for a hysteresis of -0.005 degrees on the C-axis, the generated correction component is +0.005 degrees. The physical meaning of this vector is: in order for the actual result to reach the target, we need to travel this small additional distance or angle at the command level.

[0041] Finally, and most crucially, the control system superimposes this pre-compensation correction vector onto the original multi-axis motion command sequence, forming a corrected command sequence. That is, the new X-axis command is no longer 100.000 mm, but 100.002 mm; the new C-axis command is no longer 90.000 degrees, but 90.005 degrees. This pre-exaggerated, corrected command sequence is then actually sent to the five-axis machine tool's drive system for execution. Due to the inherent physical response hysteresis of the physical machine tool (approximately -0.002 mm and -0.005 degrees), when it executes this exaggerated command, the final actual physical response will fall very precisely into the theoretical motion state we initially expected (100.000 mm and 90.000 degrees).

[0042] In this way, the entire control process forms a feedforward compensation closed loop. It doesn't try to catch up after an error occurs, but rather accurately predicts and cancels it before it happens. Returning to the scenario of machining an integral bladed disk, when the tool needs to quickly circle the blade edge, the virtual machine model anticipates that the rotary axis will rotate slightly slower due to its enormous inertia, and the linear axis will experience a slight bend due to the cantilever beam structure. Therefore, the system will pre-calculate by making the rotary axis rotate a little longer and the linear axis move a little longer, ultimately ensuring that the tool's trajectory perfectly matches the theoretical toolpath. The machined blade surface is smooth as a mirror, free from vibration marks caused by dynamic errors.

[0043] In one optional implementation, the dynamic parameters characterizing the physical properties of the mechanical transmission chain include inertial parameters, stiffness parameters, and damping parameters; the step of calling a preset virtual machine tool model and calculating the expected motion state of the end effector of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical properties of the mechanical transmission chain includes: The multi-axis motion command sequence is input into the preset virtual machine bed model. Based on the multi-axis motion command sequence, inertial parameters, stiffness parameters and damping parameters, dynamic iterative calculation is performed to obtain the expected position and expected attitude of the execution end, which are taken as the expected motion state.

[0044] Specifically, for each degree of freedom (each axis) of a machine tool, a second-order differential equation can be established to describe its motion, in the classical form of M*x''+C*x'+K*x=F(t). Here, M is the equivalent inertia parameter, C is the equivalent damping parameter, K is the equivalent stiffness parameter, x, x', and x'' are the position, velocity, and acceleration of the axis, respectively, and F(t) is the driving force or torque provided by the servo motor. When a motion command (i.e., the target position x_target) is issued, the control system calculates a driving force F(t) to propel the object towards the target. The virtual machine tool model, within each servo cycle, uses the known M, C, and K parameters and the calculated driving force F(t) to solve this differential equation using numerical integration methods (such as the Euler method or the more precise Runge-Kutta method) to obtain the actual position x for the next time step. This process is performed simultaneously for all linked axes, considering their coupling relationships, and ultimately synthesizing the expected position and orientation of the tool's end effector. The reason for performing dynamic iterative calculations is that the response of a machine tool is not instantaneous. The current state is the result of the previous state and the forces acting on it. Only through step-by-step iterative calculations can this dynamic process be realistically simulated.

[0045] In an optional implementation, after outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Real-time acquisition of output torque data from the drive system; Extract the equivalent load component related to the machining load based on the output torque data; The stiffness parameters are dynamically weakened and compensated based on the equivalent load components to correct the elastic deformation of the five-axis machine tool under load.

[0046] Modern servo drives can very accurately monitor and feedback the real-time torque output of a motor. The total torque output by the motor consists of several parts: one part is used to overcome the inertia of moving parts to achieve acceleration and deceleration, another part is used to overcome damping such as friction, and the remaining part is used to resist external loads. In machine tool machining scenarios, this external load is mainly the cutting force generated when the tool cuts the workpiece. Since the virtual machine tool model already knows the motion state (acceleration and velocity) and inertia and damping parameters of the machine tool, it can estimate the portion of torque required to overcome inertia and damping. Subtracting this estimated torque from the total torque fed back by the motor, the remaining amount can be considered as the equivalent load component directly related to the machining load. This component reflects the magnitude of the cutting force in real time. Once the equivalent load component is obtained, the system can dynamically weaken and compensate for the stiffness parameter in the virtual machine tool model. That is, when an increase in cutting load is detected, the stiffness parameter K value in the model is correspondingly lowered. In this way, when the model performs dynamic calculations, a lower K value will naturally lead to a larger displacement under the same force, i.e., a larger predicted elastic deformation. This more accurate deformation prediction will generate a more effective compensation command, thereby more precisely correcting trajectory deviations caused by cutting forces.

[0047] Furthermore, the steps for dynamically weakening and compensating for stiffness parameters based on equivalent load components include: Based on the equivalent load component, the matching stiffness attenuation coefficient is retrieved from the preset stiffness load mapping table, which records the stiffness attenuation coefficients corresponding to different loads. Multiplying the stiffness parameter by the stiffness attenuation coefficient yields the equivalent stiffness parameter of the five-axis machine tool under the current machining load, which is then used as the compensated stiffness parameter.

[0048] Specifically, calculating the relationship between load and stiffness changes in real time through complex mechanical analysis can be too time-consuming for CNC systems with limited computing power. Therefore, a stiffness load database can be established in advance through experiments or finite element simulations. This database can be simplified into a two-dimensional lookup table, namely a stiffness load mapping table.

[0049] For example, an example of a stiffness load mapping table is as follows: Equivalent load component (Nm) | Stiffness attenuation coefficient 0-10|1.00 10.1-20|0.98 20.1-30|0.95 30.1-40|0.91 40.1-50|0.86 During actual operation, the control system calculates the current equivalent load component to be 25 Nm. By consulting the table above, the system matches the range of 20.1-30 Nm, corresponding to a stiffness attenuation coefficient of 0.95. Then, the system multiplies the original stiffness parameter K_initial, calibrated under no-load conditions for that axis in the virtual machine model, with this attenuation coefficient to obtain the equivalent stiffness parameter K_effective under the current load: K_initial * 0.95. In subsequent dynamic iteration calculations, the model will use this weakened K_effective value instead of the fixed K_initial value. This table lookup and multiplication method involves minimal computation and can be completed effortlessly within each servo cycle, thus achieving efficient dynamic compensation of the stiffness parameter.

[0050] In a preferred embodiment, before outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Obtain the physical performance base information of the drive system, including the maximum torque limit of each motion axis motor, the maximum current limit of the driver, and the safe load-bearing capacity of the transmission components. Determine whether the corrected multi-axis motion command sequence reaches the physical output limit based on the physical performance base information.

[0051] The control system pre-stores a physical health check report for the machine tool, which contains information on its physical performance. This report details the core performance limits of the machine tool hardware.

[0052] For example, the physical performance base information of a specific five-axis machining center might be as follows: X-axis servo motor: maximum continuous output torque 30Nm, peak torque 90Nm.

[0053] Y-axis servo motor: maximum continuous output torque 30Nm, peak torque 90Nm.

[0054] Z-axis servo motor: maximum continuous output torque 40Nm (with brake), peak torque 120Nm.

[0055] A-axis torque motor: maximum continuous output torque 200Nm, peak torque 550Nm.

[0056] C-axis torque motor: maximum continuous output torque 180Nm, peak torque 500Nm.

[0057] Each axis driver: maximum output current 50A.

[0058] Z-axis ball screw: Maximum dynamic axial load capacity 25000N.

[0059] After generating the corrected motion command, the control system does not immediately issue it. Instead, it uses the virtual machine bed model to calculate the physical resources required to execute the corrected command. For example, based on the corrected acceleration command and the inertia parameter of the axis, the required drive torque (T=J*α) can be calculated; based on the motor torque constant, the required drive current (I=T / Kt) can be calculated. Then, the system compares these calculated requirements with the upper limits in the physical performance base information. If any requirement exceeds the limit (for example, the calculated peak torque of axis A is 580Nm, while its upper limit is 550Nm), the system determines that the corrected command has reached the physical output limit.

[0060] Furthermore, when it is determined that the modified multi-axis motion command sequence has reached the physical output limit, the step of outputting the modified multi-axis motion command sequence to the drive system includes: Calculate the output gap ratio; Based on the output gap ratio, the feed commands of all motion axes at the same moment are reduced proportionally to obtain a reduced multi-axis motion command sequence. This ensures the geometric consistency of the multi-axis linkage trajectory without exceeding the physical output limit. The reduced multi-axis motion command sequence is then output to the drive system.

[0061] The output gap ratio refers to the degree to which the corrected multi-axis motion command sequence exceeds the physical output limit. Specifically, it can be achieved by calculating the maximum percentage by which each motion axis exceeds its limit and taking the maximum value as the output gap ratio. For example, if the command torque of a certain axis is 120% of the motor's maximum torque, then the excess ratio for that axis is 20%, and the output gap ratio is 20%. Synchronously and proportionally reducing the feed commands of all motion axes at the same time means reducing the feed commands of all motion axes by the same proportion at the same point in time. Outputting the reduced multi-axis motion command sequence to the drive system means sending the adjusted command sequence to the machine tool's drive system. This can be achieved by sending the commands to the servo driver via an industrial Ethernet or fieldbus interface.

[0062] Specifically, this application aims to address the issue of modified multi-axis motion command sequences potentially reaching physical output limits, ensuring safe machine tool operation and maintaining trajectory geometric consistency. When it is determined that the modified multi-axis motion command sequence has reached the physical output limit, the output gap ratio is first calculated. This ratio reflects the degree to which the command sequence exceeds the physical limit. Next, based on this output gap ratio, the feed commands of all motion axes at the same time are simultaneously and proportionally reduced, resulting in a reduced multi-axis motion command sequence. This proportional reduction method ensures that the geometric consistency of the multi-axis linkage trajectory is maintained to the maximum extent possible without exceeding the physical output limit, avoiding overall trajectory deformation due to limitations of individual axes. Finally, the reduced multi-axis motion command sequence is output to the drive system, thereby achieving optimized control while ensuring machine tool safety and trajectory accuracy.

[0063] In one specific embodiment, the output gap ratio is first calculated. Taking the A-axis as an example, if the required torque is 580 Nm and the limit is 550 Nm, then the output gap ratio is 580 / 550 ≈ 1.055. This means that the demand exceeds the limit by 5.5%.

[0064] Next, the system will use this gap ratio as a benchmark to synchronously and proportionally reduce the feed commands of all axes involved in the linkage within the current servo cycle. The reduction ratio is the reciprocal of the gap ratio, i.e., 1 / 1.055 ≈ 0.948.

[0065] Assuming that at this moment, the corrected multi-axis motion command (taking speed as an example) is: X-axis: 200mm / s Y-axis: 150mm / s Z-axis: -50mm / s A-axis: 30 deg / s C-axis: -45deg / s The system will multiply all these speed commands by a reduction factor of 0.948 to obtain the reduced commands that are finally output to the drive system. X-axis: 200 * 0.948 = 189.6 mm / s Y-axis: 150 * 0.948 = 142.2 mm / s Z-axis: -50 * 0.948 = -47.4 mm / s A-axis: 30 * 0.948 = 28.44 deg / s C-axis: -45 * 0.948 = -42.66 deg / s In this way, the speed of all axes is reduced by the same proportion. This is equivalent to the tool moving along its original trajectory in a slightly smaller step at that instant. Although the instantaneous feed rate is reduced, the geometry (i.e., profile) of the path traced by the tool center point in space is precisely maintained because the motion ratio of all axes remains unchanged. This ensures that even in sections where hardware capabilities are at their limit, machining accuracy is not compromised, only the machining time is slightly extended.

[0066] Furthermore, while the virtual machine tool model is highly accurate during initial calibration, its physical properties may slowly change over long-term operation of the physical machine tool. For example, bearing wear can alter frictional characteristics, or prolonged operation can cause thermal deformation due to structural temperature rise. These changes lead to a gradual mismatch between the virtual machine tool model and the physical machine tool, a phenomenon known as parameter drift. If left unaddressed, the model's predictions will become increasingly inaccurate, and the effectiveness of feedforward compensation will be significantly reduced.

[0067] In a preferred embodiment, after outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Real-time acquisition of the actual position feedback of the drive system; The actual position feedback is compared with the expected motion state to determine the position deviation characteristics; When the position deviation characteristics meet the preset trend increase condition, it is determined that the physical model parameters of the five-axis machine tool have drifted, and the model parameters of the preset virtual machine tool model are adjusted based on the position deviation characteristics so that they are refitted to the current physical model parameters of the five-axis machine tool.

[0068] Specifically, during operation, the control system simultaneously possesses three sets of data: theoretical motion state (from the original command), expected motion state (from the prediction of the virtual machine tool model), and actual position feedback (from the actual readings of the physical machine tool encoder). Ideally, after feedforward compensation, the actual position feedback should be highly consistent with the theoretical motion state. This step focuses on the deviation between the actual position feedback and the expected motion state, which reflects the prediction accuracy of the virtual machine tool model.

[0069] The system continuously monitors this prediction bias. A small and random bias is normal, but if the bias exhibits a regular and continuously increasing trend, it indicates a problem with the model. For example, the system might detect that in the last half hour of processing, the actual position of the physical machine tool consistently lags behind the position predicted by the virtual machine tool model by 0.001 millimeters, and this lag is slowly increasing. This is a typical example of a trending increase. This condition can be quantified; for instance, if the moving average of the prediction bias increases by more than 5% compared to the previous statistical period over 1000 consecutive servo cycles, then parameter drift is detected.

[0070] Once parameter drift is detected, the system needs to fine-tune the parameters of the virtual machine bed model.

[0071] The steps for adjusting the model parameters of the preset virtual machine bed model based on position deviation characteristics include: The changes in the physical model parameters of the five-axis machine tool are calculated in reverse based on the characteristics of the positional deviation. The damping parameter or discrete time step integral parameter in the preset virtual machine bed model is adjusted in real time according to the changes in the physical model parameters.

[0072] This step is the specific execution action of model self-calibration. The system analyzes the characteristics of the prediction bias and correlates them with changes in specific physical parameters.

[0073] For example, if the overshoot and oscillations during acceleration and deceleration are more pronounced during the acceleration and deceleration phases compared to the expected position, this usually means that the actual damping of the system has decreased (e.g., due to deteriorated lubricating oil or wear leading to reduced friction). In this case, the system will work backward to calculate how much adjustment is needed to the damping parameters in the model to reproduce this overshoot, and then adjust the damping parameters in the model accordingly.

[0074] If the actual position consistently lags behind the expected position systematically, and the lag is proportional to the velocity, this may indicate a change in the viscous friction coefficient. The system will then adjust the damping parameters in the model accordingly.

[0075] If the deviation manifests as a minor timing misalignment, it may be related to delays in the system's internal data processing. In this case, the discrete time step integration parameters in the virtual machine bed model's numerical integrator can be fine-tuned to realign the model with the real-world time axis.

[0076] Through this online, continuous fine-tuning, the virtual machine tool model can dynamically track the performance evolution of the physical machine tool, always maintaining its accuracy as a digital twin, thereby ensuring that the feedforward compensation strategy remains efficient and stable throughout the entire lifecycle of the machine tool.

[0077] Furthermore, the steps of comparing the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis include: Calculate the physical response hysteresis deviations for the linear axis and the rotation axis respectively; For a linear axis, calculate the position lag deviation vector value between the expected position of the expected motion state and the theoretical position of the theoretical motion; For the rotation axis, calculate the attitude angle hysteresis vector value between the expected attitude angle of the expected motion state and the theoretical attitude angle of the theoretical motion.

[0078] Specifically, for linear axes like X, Y, and Z, their positions are described in a three-dimensional Cartesian coordinate system. Therefore, calculating the deviation is very intuitive; it's simply a subtraction between two position vectors. If the theoretical position vector is P_theory=(x_t,y_t,z_t), and the model-predicted expected position vector is P_predict=(x_p,y_p,z_p), then the position lag deviation vector value is ΔP=P_theory-P_predict=(x_t-x_p,y_t-y_p,z_t-z_p). This vector directly gives the direction and magnitude of the compensation needed in space.

[0079] For rotation axes like A, B, and C, the situation is more complex. Directly subtracting angle values ​​encounters periodicity issues (e.g., 359 degrees differs from 1 degree by only 2 degrees, but direct subtraction yields 358 degrees). A more rigorous approach is to use mathematical tools that uniquely describe spatial attitude, such as rotation matrices or quaternions. For example, theoretical attitude angles (e.g., A_t, C_t) and expected attitude angles (e.g., A_p, C_p) can be converted into corresponding rotation matrices R_theory and R_predict, respectively. Then, the error rotation matrix R_error = R_theory * (R_predict)^-1 required to transform from the expected attitude to the theoretical attitude is calculated, where (R_predict)^-1 is the inverse of R_predict. Finally, the corresponding attitude angle hysteresis vector value (e.g., represented as a tiny rotation axis and rotation angle) is inversely solved from this error rotation matrix R_error. This vector value accurately describes the attitude deviation, and subsequent compensation is based on this vector, thus ensuring the correctness of the rotation compensation.

[0080] Secondly, see Figure 2 This application also provides a multi-axis linkage trajectory optimization control system for a five-axis linkage machining center, the system comprising: The instruction acquisition module 210 is used to acquire the multi-axis motion instruction sequence of a five-axis machine tool; The state prediction module 220 is used to call a preset virtual machine tool model and calculate the expected motion state of the execution end of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical characteristics of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. The deviation determination module 230 is used to compare the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. Vector generation module 240 is used to generate a corresponding pre-compensation correction vector based on the physical response hysteresis deviation. The instruction output module 250 is used to correct the multi-axis motion instruction sequence based on the pre-compensation correction vector and output the corrected multi-axis motion instruction sequence to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion instruction sequence is synchronized with the theoretical motion state. The stiffness parameter compensation module is used to acquire the output torque data of the drive system in real time after the corrected multi-axis motion command sequence is output to the drive system of the five-axis machine tool. Extract the equivalent load component related to the machining load based on the output torque data; The stiffness parameters are dynamically weakened and compensated based on the equivalent load components to correct the elastic deformation of the five-axis machine tool under load. The physical limit verification module is used to check the corrected multi-axis motion command sequence before outputting it to the drive system of the five-axis machine tool. Obtain the physical performance base information of the drive system, including the maximum torque limit of each motion axis motor, the maximum current limit of the driver, and the safe load-bearing capacity of the transmission components. Determine whether the corrected multi-axis motion command sequence reaches the physical output limit based on the physical performance base information; The model adjustment module is used after outputting the corrected multi-axis motion command sequence to the drive system of the five-axis machine tool. Real-time acquisition of the actual position feedback of the drive system; The actual position feedback is compared with the expected motion state to determine the position deviation characteristics; When the position deviation characteristics meet the preset trend increase condition, it is determined that the physical model parameters of the five-axis machine tool have drifted, and the model parameters of the preset virtual machine tool model are adjusted based on the position deviation characteristics so that they are refitted to the current physical model parameters of the five-axis machine tool.

[0081] By modularizing the functional division, the various components required to achieve dynamic error feedforward compensation and their collaborative relationships are clearly defined, providing a clear system-level blueprint for the engineering implementation of this method.

[0082] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center, characterized in that, The method includes: Obtain the multi-axis motion command sequence of a five-axis machine tool; A preset virtual machine tool model is invoked, and the expected motion state of the execution end of the five-axis machine tool is calculated based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical properties of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. The expected motion state is compared with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. Generate a corresponding pre-compensation correction vector based on the physical response hysteresis deviation; The multi-axis motion command sequence is corrected based on the pre-compensation correction vector, and the corrected multi-axis motion command sequence is output to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion command sequence is synchronized with the theoretical motion state. Before outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Obtain the physical performance base information of the drive system, which includes the maximum torque limit of each motion axis motor, the maximum current limit of the driver, and the safe load-bearing capacity of the transmission components. Based on the physical performance base information, determine whether the corrected multi-axis motion command sequence has reached the physical output limit; When it is determined that the modified multi-axis motion command sequence has reached the physical output limit, the step of outputting the modified multi-axis motion command sequence to the drive system includes: Calculate the output gap ratio; Based on the output gap ratio, the feed commands of all motion axes at the same time are synchronously and proportionally reduced to obtain the reduced multi-axis motion command sequence, so as to maintain the geometric consistency of the multi-axis linkage trajectory without exceeding the physical output limit, and the reduced multi-axis motion command sequence is output to the drive system.

2. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 1, characterized in that, The dynamic parameters characterizing the physical properties of the mechanical transmission chain include inertia parameters, stiffness parameters, and damping parameters; the step of calling a preset virtual machine tool model and calculating the expected motion state of the execution end of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical properties of the mechanical transmission chain includes: The multi-axis motion command sequence is input into the preset virtual machine bed model. Dynamic iterative calculations are performed based on the multi-axis motion command sequence, the inertial parameters, the stiffness parameters, and the damping parameters to obtain the expected position and expected attitude of the execution end, which are taken as the expected motion state.

3. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 2, characterized in that, After outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Real-time acquisition of the output torque data of the drive system; Extract the equivalent load component related to the machining load based on the output torque data; The stiffness parameter is dynamically weakened and compensated based on the equivalent load component to correct the elastic deformation of the five-axis machine tool under load.

4. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 3, characterized in that, The step of dynamically weakening and compensating the stiffness parameter based on the equivalent load component includes: Based on the equivalent load component, a matching stiffness attenuation coefficient is retrieved from a preset stiffness load mapping table, wherein the preset stiffness load mapping table records the stiffness attenuation coefficients corresponding to different loads. Multiplying the stiffness parameter by the stiffness attenuation coefficient yields the equivalent stiffness parameter of the five-axis machine tool under the current machining load, which is then used as the compensated stiffness parameter.

5. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 1, characterized in that, After outputting the modified multi-axis motion command sequence to the drive system of the five-axis machine tool, the method further includes: Real-time acquisition of the actual position feedback of the drive system; The actual position feedback is compared with the expected motion state to determine the position deviation characteristics; When the position deviation feature meets the preset trend increase condition, it is determined that the physical model parameters of the five-axis machine tool have drifted, and the model parameters of the preset virtual machine tool model are adjusted based on the position deviation feature so that they are refitted to the current physical model parameters of the five-axis machine tool.

6. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 5, characterized in that, The step of adjusting the model parameters of the preset virtual machine bed model based on the position deviation characteristics includes: The changes in the physical model parameters of the five-axis machine tool are calculated in reverse based on the position deviation characteristics. The damping parameter or discrete time step integral parameter in the preset virtual machine bed model is adjusted in real time according to the changes in the physical model parameters.

7. The method for optimizing and controlling the multi-axis linkage trajectory of a five-axis linkage machining center according to claim 1, characterized in that, The step of comparing the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis includes: Calculate the physical response hysteresis deviations for the linear axis and the rotation axis respectively; For the linear axis, calculate the position lag deviation vector value between the expected position of the expected motion state and the theoretical position of the theoretical motion; For the rotation axis, calculate the attitude angle hysteresis deviation vector value between the expected attitude angle of the expected motion state and the theoretical attitude angle of the theoretical motion.

8. A multi-axis linkage trajectory optimization control system for a five-axis linkage machining center, used to execute the multi-axis linkage trajectory optimization control method for a five-axis linkage machining center as described in any one of claims 1 to 7, characterized in that, The system includes: The instruction acquisition module is used to acquire the multi-axis motion instruction sequence of a five-axis machine tool; The state prediction module is used to call a preset virtual machine tool model and calculate the expected motion state of the execution end of the five-axis machine tool based on the multi-axis motion command sequence and the dynamic parameters characterizing the physical characteristics of the mechanical transmission chain. The running cycle of the preset virtual machine tool model is consistent with the servo control cycle of the five-axis machine tool, and the model parameters of the preset virtual machine tool model are consistent with the physical model parameters of the five-axis machine tool. The deviation determination module is used to compare the expected motion state with the theoretical motion state corresponding to the multi-axis motion command sequence to determine the physical response hysteresis deviation of each motion axis. The vector generation module is used to generate a corresponding pre-compensation correction vector based on the physical response hysteresis deviation. The instruction output module is used to correct the multi-axis motion instruction sequence based on the pre-compensation correction vector, and output the corrected multi-axis motion instruction sequence to the drive system of the five-axis machine tool so that the physical response state of the five-axis machine tool when executing the corrected multi-axis motion instruction sequence is synchronized with the theoretical motion state.

Citation Information

Patent Citations

  • Numerical control machine tool contour error prediction modeling method and device, equipment and storage medium

    CN118627224A

  • Bogie numerical control machining error compensation method and system based on digital twinning

    CN120630875A

  • Method and system for predicting contour error of digital twin five-axis machine tool and electronic equipment

    CN121578742A

  • Feed speed control method for a numerical control device

    US5200680A