Error compensation method for double-drive synchronization system of gantry machining center

Abstract

The application discloses a gantry machining center double-drive synchronous system error compensation method and relates to the technical field of double-drive synchronous error compensation.The gantry machining center double-drive synchronous system error compensation method comprises the following steps: the synchronous error of the double-drive system of the gantry machining center is divided into two physical parameters, namely a linear offset and a curved deformation; l is the linear offset, which represents the average position deviation of two driving shafts in the ideal motion direction of the double-drive system of the gantry machining center and reflects the overall positioning inaccuracy; and w is the curved deformation, which reflects the angular distortion degree of the gantry beam of the gantry machining center caused by uneven force or thermal deformation.The gantry machining center double-drive synchronous system error compensation method is an error compensation method based on dynamic characteristic modeling, a synchronous error model is constructed by introducing the linear offset l and the curved deformation w, and high-precision compensation is realized by combining a target function optimization strategy, so that the synchronous precision is improved.

Description

Technical Field

[0001] This invention relates to the field of dual-drive synchronous error compensation technology, specifically to a method for error compensation in a dual-drive synchronous system of a gantry machining center. Background Technology

[0002] In CNC machine tools, gantry machining centers, due to their large span and high load characteristics, often adopt a dual servo drive (dual drive) system to improve motion stability and load-bearing capacity.

[0003] However, due to factors such as mechanical structure asymmetry, thermal deformation, transmission backlash, and servo response differences, the drive shafts on both sides of a gantry machining center are prone to synchronization errors, leading to beam distortion or misalignment, which severely affects machining accuracy. Therefore, this invention proposes an error compensation method for the dual-drive synchronization system of a gantry machining center to solve this problem. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an error compensation method for a dual-drive synchronous system in a gantry machining center. This method, based on dynamic characteristic modeling, introduces linear offset. With curvature deformation A synchronization error model was constructed, and a high-precision compensation was achieved by combining it with an objective function optimization strategy, thus solving the defects and shortcomings of the existing technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for error compensation in a dual-drive synchronous system of a gantry machining center includes:

[0007] The synchronization error of the dual-drive system of the gantry machining center is divided into two physical parameters: linear offset and curvilinear deformation.

[0008] set up The linear offset represents the average positional deviation of the two drive shafts in the ideal motion direction in the dual-drive system of the gantry machining center. It reflects the overall positioning inaccuracy problem, such as the overall displacement deviation caused by guide rail installation error, encoder zero point offset, etc.

[0009] set up The angular deformation reflects the degree of angular torsion of the gantry beam in the gantry machining center caused by uneven stress or thermal deformation, i.e. the "warping" effect caused by asynchronous driving on both sides, and is measured in micrometers or millimeters.

[0010] Let δ be the angle between the linear offset and the curvilinear deformation, which represents the spatial orientation angle of the principal axis of deformation relative to the direction of motion;

[0011] In the tangential coordinate system along the actual motion trajectory, and Projecting onto the x and y axes yields the combined error components:

[0012] Δx= + cosδ;

[0013] Δy= sinδ;

[0014] Among them, Δx is the total error in the x-direction, which includes positioning deviation and partial torsional deformation; Δy is the lateral offset error in the y-direction, which is entirely caused by structural deformation and is an important indicator for judging synchronous imbalance.

[0015] This projection model transforms complex multi-source errors into a quantifiable and controllable two-dimensional vector space representation, facilitating subsequent closed-loop compensation design.

[0016] To effectively suppress the aforementioned errors, an objective function J is defined as the minimum distance metric between the current error state and the desired state of the dual-drive system of the gantry machining center. A reference datum C is defined. 1 x and C 1y C 1 x Let C be the ideal compensation target value in the x-direction. 1y Let y be the ideal compensation target value. Setting it to 0 indicates no residual error. Construct the error square term:

[0017] T = (ΔxC) 1 x ) 2 + (Δy-C) 1y ) 2 ;

[0018] make , representing the Euclidean distance from the current error vector to the target point;

[0019] Introducing an ideal radius R, we construct an objective function in the form of a modulo operation:

[0020] J = |R - t|;

[0021] In the above formula, as t approaches R, J approaches 0, at which point the dual-drive system of the gantry machining center is in the "tolerance error loop". If R is set as a minimum constant and ε is the threshold, when J < ε, the system is considered to have entered the high-precision synchronization zone. By adjusting the controller output, J is continuously reduced, achieving infinite approximation of the threshold ε during the dynamic process, i.e., J ≤ ε. Here, ε is not a single value, but refers to the entire convergence condition range, often represented in the form of a tolerance band in the control system.

[0022] Preferably, the following steps are performed in each control cycle of the compensation algorithm:

[0023] Step 1, Real-time Data Collection:

[0024] Obtain the actual position feedback x1 and x2 of the left and right servo axes of the dual-drive system of the gantry machining center, and calculate the average position: x m = (x1 + x2) / 2, difference: d = x1 - x2;

[0025] Step 2, Identification and :

[0026] Assumption = x m - x ideal , where x ideal The positional error of the ideal trajectory;

[0027] ∝ |d| / 2, where |d| / 2 represents the differential deformation.

[0028] Fitting angle δ using historical data;

[0029] Step 3, calculate Δx and Δy:

[0030] The projection of the error onto the local coordinate system is obtained using trigonometric relationships;

[0031] Step 4, update T, t, J

[0032] Calculate the current objective function value and evaluate whether J≤ε is satisfied;

[0033] Step 5, Generate compensation instructions

[0034] If J > ε, then the speed and position commands of the two axes are adjusted through feedforward and feedback methods:

[0035] right For those that are too large, apply translational correction;

[0036] right The resulting Δy anomaly is addressed by implementing reverse torque ratio adjustment to suppress the torsional trend.

[0037] Preferably, the step of fitting the angle δ using historical data is as follows:

[0038] 1) Data preprocessing

[0039] Real-time data acquisition: x1 and x2 are acquired in each control cycle, and calculations are performed. and ;

[0040] Historical data cache: stores data from the past N periods. , Data pairs are used to form a sample set;

[0041] 2) Model selection

[0042] The estimated value of δ is fitted using linear regression or least squares method: = δ* +P, where P is a noise term that can be ignored. The optimal δ is obtained by minimizing the sum of squared residuals:

[0043] ;

[0044] The mean of historical differential deformation: This reflects the long-term statistical characteristics of the deformation. This represents the mean of historical positioning errors: This reflects the long-term statistical characteristics of systematic errors.

[0045] This invention provides an error compensation method for a dual-drive synchronous system in a gantry machining center. It has the following beneficial effects:

[0046] The error compensation method for the dual-drive synchronous system of the gantry machining center of the present invention divides the dynamic characteristics of the dual-drive synchronous system into linear offset. and curvature deformation Two parameter values, and The angle between them is Given an objective function J, a dynamic model is established by setting a threshold range that is as small as possible. Based on the given objective function, a compensation algorithm is established. By infinitely approximating the threshold, the dual-drive synchronization system compensates for the positioning errors of the two synchronous shafts and the asynchrony errors of the two shafts. This can compensate for the errors of the dual-drive synchronization system under different working conditions and improve the synchronization accuracy. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the execution steps within each control cycle of the compensation algorithm of this invention;

[0048] Figure 2 This is a flowchart illustrating step 5 of the present invention;

[0049] Figure 3 This is a schematic diagram of the linear offset and curvilinear deformation of the present invention;

[0050] Figure 4 This is a schematic block diagram showing the linear offset, curvilinear deformation, and objective function of the present invention.

[0051] Figure 5 This is a schematic diagram of an existing dual-drive gantry.

[0052] Attached reference numerals: 1. Dual drive shaft one; 2. Dual drive shaft two; 3. Crossbeam; 4. Spindle. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Example:

[0055] like Figure 1-5 As shown, this embodiment of the invention provides an error compensation method for a dual-drive synchronous system of a gantry machining center. Based on the multi-servo drive system of the gantry machining center, it can compensate for the error of the dual-drive synchronous system of the gantry machining center under different working conditions, thereby improving the synchronization accuracy.

[0056] The error compensation method for the dual-drive synchronous system of the gantry machining center includes:

[0057] The synchronization error of the dual-drive system of the gantry machining center is divided into two physical parameters: linear offset and curvilinear deformation.

[0058] set up The linear offset represents the average positional deviation of the two drive shafts in the ideal motion direction in the dual-drive system of the gantry machining center. It reflects the overall positioning inaccuracy problem, such as the overall displacement deviation caused by guide rail installation error, encoder zero point offset, etc.

[0059] set up The angular deformation reflects the degree of angular torsion of the gantry beam in the gantry machining center caused by uneven stress or thermal deformation, i.e. the "warping" effect caused by asynchronous driving on both sides, and is measured in micrometers or millimeters.

[0060] Let δ be the angle between the linear offset and the curvilinear deformation, which represents the spatial orientation angle of the principal axis of deformation relative to the direction of motion;

[0061] In the tangential coordinate system along the actual motion trajectory, and Projecting onto the x and y axes yields the combined error components:

[0062] Δx= + cosδ;

[0063] Δy= sinδ;

[0064] Among them, Δx is the total error in the x-direction, which includes positioning deviation and partial torsional deformation; Δy is the lateral offset error in the y-direction, which is entirely caused by structural deformation and is an important indicator for judging synchronous imbalance.

[0065] This projection model transforms complex multi-source errors into a quantifiable and controllable two-dimensional vector space representation, facilitating subsequent closed-loop compensation design.

[0066] To effectively suppress the aforementioned errors, an objective function J is defined as the minimum distance metric between the current error state and the desired state of the dual-drive system of the gantry machining center. A reference datum C is defined. 1 x and C 1y C 1 x Let C be the ideal compensation target value in the x-direction. 1y Let y be the ideal compensation target value. Setting it to 0 indicates no residual error. Construct the error square term:

[0067] T = (ΔxC) 1 x ) 2 + (Δy-C) 1y ) 2 ;

[0068] make , representing the Euclidean distance from the current error vector to the target point;

[0069] Introducing an ideal radius R, we construct an objective function in the form of a modulo operation:

[0070] J = |R - t|;

[0071] In the above formula, as t approaches R, J approaches 0, at which point the dual-drive system of the gantry machining center is in the "tolerance error loop". If R is set as a minimum constant and ε is the threshold, when J < ε, the system is considered to have entered the high-precision synchronization zone. By adjusting the controller output, J is continuously reduced, achieving infinite approximation of the threshold ε during the dynamic process, i.e., J ≤ ε. Here, ε is not a single value, but refers to the entire convergence condition range, often represented in the form of a tolerance band in the control system.

[0072] In this embodiment, the determination of the ideal radius R typically considers multiple factors, including processing requirements, inherent system characteristics, experimental calibration, and control feasibility, as detailed below:

[0073] 1) Based on machining process accuracy requirements: According to the part machining tasks undertaken by the gantry machining center, extract core accuracy indicators such as form and position tolerances and dimensional tolerances, and decompose them into the maximum allowable synchronization error threshold of the dual-drive synchronous system. This threshold is used as the basic reference value for R. For example, if the flatness requirement of the part is 0.02mm, then R needs to be set to be no greater than the synchronization error component corresponding to this tolerance, ensuring that the machining accuracy is met after the system error is suppressed.

[0074] 2) Calibration based on inherent error characteristics of the system: By conducting error tests on the dual-drive synchronous system under no-load and load conditions, data such as the position deviation of the left and right servo axes and the deformation of the crossbeam are collected. The distribution range and peak value of the error are analyzed, and the statistically significant extreme value, such as the upper limit of the error in the 95% confidence interval, is taken as the basis for the value of R. This ensures that R conforms to the actual error performance of the system and avoids the target setting from deviating from the hardware capability.

[0075] 3) Considering the convergence feasibility of the control algorithm: Taking into account the controller's response bandwidth, adjustment speed, and other performance characteristics, if R is set too small, the controller may cause system oscillations due to over-adjustment; if it is too large, it will not meet the high precision requirements. Therefore, it is necessary to determine the error range within which the controller can converge stably through simulation or pre-experimentation, and set R at the balance point between control stability and accuracy requirements.

[0076] 4) Adaptive adjustment in dynamic scenarios: For different processing loads such as heavy cutting and light finishing, multiple R values ​​can be preset, and the system can automatically switch according to the real-time working conditions, which can ensure the control stability during heavy cutting and meet the high precision requirements during finishing.

[0077] Furthermore, the "tolerable error loop" refers to the error closed-loop range constructed with the ideal compensation target as the center and the ideal radius R as the boundary. It is a basic quantitative definition of the system synchronization error state. If there is no residual error, it is set to 0. As the basic judgment benchmark for the error state of the dual-drive synchronization system, when the objective function J equals R, it means that the system synchronization error is just within the preset basic qualified boundary, neither exceeding the allowable error range nor entering a higher precision range. This provides a basic reference for synchronization control and ensures the basic precision requirements of processing.

[0078] Furthermore, the high-precision synchronization zone refers to a more stringent error range defined within the allowable error loop, bounded by a minimal constant threshold ε, corresponding to the high-precision machining requirements of gantry machining centers. As a precision judgment standard for high-end machining scenarios, when the objective function J is less than ε, it indicates that the positioning error and asynchronous error of the dual-drive two-axis system have been suppressed to a minimum, meeting the extremely high precision requirements of aerospace, precision molds, and other applications. This is a core indicator that the system's synchronization accuracy has reached an advanced level. The value of the high-precision synchronization zone threshold ε follows the core logic of "demand-oriented, hardware adaptation, and experimental calibration."

[0079] 1) Based on the requirements, using the form and position tolerances of high-end parts such as aerospace and precision molds as the benchmark, extract the allowable component of synchronous error, take 1 / 3 to 1 / 2 of the corresponding tolerance of the part as the initial value, and reserve accuracy redundancy;

[0080] 2) Through hardware adaptation, the minimum controllable error of the machine tool dual-drive system can be measured using a laser interferometer. The minimum controllable error referred to here includes the transmission chain accuracy and the sensor resolution limit. ε is set to 1.1 to 1.2 times this value to avoid exceeding the hardware capability and causing system oscillation.

[0081] 3) In engineering practice, values ​​can be obtained through experimental calibration. For example, compensation experiments can be carried out under high-precision working conditions such as no-load and high-speed precision repair, and ε can be finely adjusted step by step until the synchronization error is stable and converged and the system has no abnormal fluctuations, and finally the optimal threshold can be determined.

[0082] Furthermore, in this embodiment, the tolerance band is an abbreviation for the allowable error fluctuation range or range, not a single fixed value, but a dynamic range of conditions for error convergence. Its main function is to provide a flexible and stable operating space for the controller's error adjustment, avoiding frequent system adjustments and oscillations caused by pursuing absolute zero error, while ensuring that the synchronization accuracy is always within an acceptable fluctuation range, adapting to dynamic working conditions such as load changes and temperature fluctuations during processing, and balancing the stability and accuracy requirements of the control.

[0083] Specifically, the selection of the tolerance zone needs to be determined comprehensively based on the machining accuracy requirements, the inherent characteristics of the system, experimental verification, and dynamic operating conditions, mainly considering the following conditions:

[0084] 1) If the primary goal is to achieve high machining accuracy, the basic tolerance range needs to be determined based on the industry accuracy standards of the workpiece. For example, in the aerospace field, where the accuracy requirements of parts are extremely high, the tolerance range for precision components such as aircraft wing boxes and engine parts should be set in the range of 0.005 to 0.01 mm. For the machining of ordinary heavy machinery parts, the tolerance range can be widened to 0.02 to 0.05 mm to ensure that the accuracy of the parts after compensation meets the tolerance requirements of the drawings.

[0085] 2) Considering the inherent error characteristics of the matching system, the inherent errors of the gantry machining center can be measured by using equipment such as laser interferometers and displacement sensors. These errors mainly include mechanical transmission clearance, thermal deformation fluctuation value, and static synchronization error of dual drive shafts. The tolerance band should be 1.2 to 1.5 times greater than the minimum uncontrollable error of the system to avoid the controller from frequently adjusting and oscillating due to pursuing accuracy beyond the hardware capability.

[0086] 3) Combined with dynamic working condition adaptive adjustment, it is necessary to set different tolerance bands for different scenarios in response to dynamic factors such as load changes and temperature fluctuations during the machining process. For example, during no-load debugging or fine finishing processes, the tolerance band takes the minimum value; under heavy cutting and high-speed feed conditions, due to the increase in cutting force and thermal deformation error, the tolerance band is temporarily enlarged by 20% to 30% to adapt to the fluctuation range of dynamic error, balance control stability and accuracy, and can be used with adaptive compensation algorithms to achieve precise adjustment.

[0087] 4) Optimize values ​​through experimental calibration. This mainly refers to conducting multi-condition compensation experiments on actual machine tools under feasible experimental conditions. This includes setting multiple sets of tolerance zone candidate values, testing synchronization accuracy, system response speed and oscillation under different values, and selecting the smallest interval where the synchronization accuracy meets the requirements and the system has no obvious oscillation as the final tolerance zone. At the same time, the parameter settings of mature equipment of the same model can be referenced to shorten the calibration cycle.

[0088] In a preferred embodiment of this invention, the following steps are performed in each control cycle of the compensation algorithm:

[0089] Step 1, Real-time Data Collection:

[0090] Obtain the actual position feedback x1 and x2 of the left and right servo axes of the dual-drive system of the gantry machining center, and calculate the average position: x m = (x1 + x2) / 2, difference: d = x1 - x2;

[0091] Step 2, Identification and :

[0092] Assumption = x m - x ideal , where x ideal The positional error of the ideal trajectory;

[0093] ∝ |d| / 2, where |d| / 2 represents the differential deformation.

[0094] Fitting angle δ using historical data;

[0095] Step 3, calculate Δx and Δy:

[0096] The projection of the error onto the local coordinate system is obtained using trigonometric relationships;

[0097] The triangular relationship is as follows:

[0098] Δx= + cosδ;

[0099] Δy= sinδ;

[0100] Step 4, update T, t, J

[0101] Calculate the current objective function value and evaluate whether J≤ε is satisfied;

[0102] Step 5, Generate compensation instructions

[0103] If J > ε, then the speed and position commands of the two axes are adjusted through feedforward and feedback methods:

[0104] right For those that are too large, apply translational correction;

[0105] right The resulting Δy anomaly is addressed by implementing reverse torque ratio adjustment to suppress the torsional trend.

[0106] In this embodiment, (Ideal trajectory position error) is the theoretical position reference of a single servo axis in a dual-drive synchronous system under ideal working conditions: that is, assuming no mechanical wear, no load disturbance, no control delay, no thermal deformation, or other error factors, the precise position coordinates that the servo axis should reach at each moment when running along the preset machining trajectory. It is the "perfect reference line" for measuring axis positioning offset, relative to the actual position. The difference, i.e., the above. It is used to quantify the positioning error of a single axis and is one of the basic parameters for synchronous error compensation in a dual-drive system.

[0107] Furthermore, The method for obtaining (ideal trajectory position error) is as follows:

[0108] 1) Extraction via G-code parsing: This involves parsing the target position command sequence corresponding to the servo axis from the machining G-code generated by the CAM software and directly converting it into the time-varying parameters. ;

[0109] 2) Obtained through simulation using an ideal dynamic model, that is, based on the design parameters of the gantry machining center, such as the lead screw pitch, guide rail accuracy, and motor characteristics, a dynamic simulation model is built under the assumption of no error to simulate the motion process of the axis and output the ideal position curve;

[0110] 3) Obtained through high-precision calibration. A laser interferometer can be used to perform full-stroke accuracy calibration on the machine tool. The error-free reference trajectory measured under no-load and standard temperature and humidity conditions serves as... The measured calibration value;

[0111] 4) For new machine tools, the theoretical position can be calculated using relevant design parameters, such as lead screw and pulse equivalent. The theoretical position is calculated by combining motion commands and then determined after initial accuracy calibration.

[0112] As a preferred embodiment of this invention, the step of fitting the angle δ using historical data is as follows:

[0113] 1) Data preprocessing

[0114] Real-time data acquisition: x1 and x2 are acquired in each control cycle, and calculations are performed. and ;

[0115] Historical data cache: stores data from the past N periods. , Data pairs are used to form a sample set;

[0116] 2) Model selection

[0117] The estimated value of δ is fitted using linear regression or least squares method: = δ* +P, where P is a noise term that can be ignored. The optimal δ is obtained by minimizing the sum of squared residuals:

[0118] ;

[0119] in The mean of historical differential deformation: This reflects the long-term statistical characteristics of the deformation. This represents the mean of historical positioning errors: This reflects the long-term statistical characteristics of systematic errors.

[0120] In this embodiment, the control cycle is not the mechanical cycle of the reciprocating drive of the gantry machining center, but a fixed time interval for the CNC system, including the servo controller, to execute a complete control process. It is a real-time scheduling benchmark at the software level and is completely independent of the mechanical motion cycle, such as the motion cycle of the gantry beam from the left end to the right end and back, which is usually a process of tens of seconds to several minutes. Typical values ​​are in the millisecond range, such as 1ms or 2ms.

[0121] The physical significance of the control cycle related to a dual-drive gantry milling machine mainly includes the following aspects:

[0122] 1) Real-time cycle of synchronous compensation: This is the execution unit of the error compensation algorithm. It completes position feedback acquisition, error calculation, and compensation command output once in each cycle to ensure real-time suppression of dynamic synchronization error and avoid amplification of error due to control lag.

[0123] 2) Matching the servo system response characteristics: Coordinating with the current loop and speed loop cycles of the servo axis to ensure that control commands can be responded to by the servo system in a timely manner, balancing control accuracy and system stability;

[0124] 3) The time dimension basis of error modeling: The “N periods” of historical data caching are error sample sets of N consecutive millisecond-level time nodes, which provide continuous time series data support for fitting dynamic parameters such as differential deformation and angle, and realize accurate identification of time-varying errors.

[0125] As a preferred embodiment of this invention, step 5, which generates the compensation instruction in each control cycle of the compensation algorithm described above, is specifically implemented as follows:

[0126] 1) Generation of combined feedforward and feedback compensation commands:

[0127] Feedforward compensation is based on error parameters identified in real time. , , The theoretical compensation amount is calculated in advance to offset the modelable systematic deviations, such as mechanical installation errors and static deformation.

[0128] Feedback compensation is based on the dynamic deviation between the objective function J and the threshold ε. It uses closed-loop control to suppress unmodeled disturbances, such as sudden load changes or high-frequency vibrations. The specific implementation steps are as follows:

[0129] (1) Perform feedforward term calculation:

[0130] From the formula Correct the overall positioning deviation The translation compensation amount can be obtained from the formula. To suppress lateral drift The torsional deformation compensation amount is obtained. Among them, K p This refers to the positional stiffness gain, which is related to the stiffness of the mechanical structure. Torsional stiffness gain refers to the equivalent torsional stiffness of a beam.

[0131] (2) Feedback item generation:

[0132] set up for The gradient determines the compensation direction, and the error feedback quantity can be expressed as:

[0133] ;

[0134] Among them, K fb The feedback gain coefficient controls the closed-loop convergence speed.

[0135] (3) Instruction synthesis:

[0136] The commands output to the dual-axis servo system can be obtained from the above steps, summarized as follows:

[0137] ;

[0138] In the formula, i=1,2 refer to dual drive shaft 1 and dual drive shaft 2, respectively;

[0139] 2) To Translation correction for excessively large values:

[0140] Linear offset To reflect the overall lag or lead of the two axes, it is necessary to apply reverse displacement compensation to the axis with the larger error to adjust the average position. Approaching the ideal trajectory The specific implementation steps are as follows:

[0141] (1) Identify the dominant axis of deviation:

[0142] Calculate the uniaxial error, expressed as: ,choose A larger axis is used as the compensation object, for example when > In this case, the dual drive shaft 1 is corrected; specifically, the error of each shaft is first calculated, i.e. and The combined writing is ,when > When, then correct the dual drive shaft -1, when < When that happens, then the dual drive shaft 2 is corrected.

[0143] (2) Generation of translation commands:

[0144] Compensation amount according to Calculations are performed here, if it involves lag axis acceleration compensation, then... ; Advance axle reduction compensation, then is ;

[0145] In the formula, To dynamically compensate for displacement. The feedback gain coefficient controls the closed-loop convergence speed. This is a sign function. By determining the direction of the error, i.e., whether it is lagging or leading, the polarity of the compensation amount is determined, i.e., acceleration or deceleration, ensuring that the compensation command is opposite to the direction of the error.

[0146] Its mathematical definition is:

[0147]

[0148] (3) Based on the above steps, correct the position command:

[0149] Represented as ;

[0150] In the formula, To correct the instruction position, The reference position is the target position of the ideal motion trajectory generated by the user or the upper-level controller, representing the theoretical position that the servo axis should reach under error-free conditions.

[0151] 3) If Δy is abnormal, the reverse torque ratio should be adjusted as follows:

[0152] Lateral drift Caused by torsional deformation of the crossbeam, a reverse corrective torque needs to be generated through differentiated torque output to suppress the "warping" effect.

[0153] The implementation steps are as follows:

[0154] (1) Diagnostic method for deformity:

[0155] like , As a threshold, based on Determine the direction of the twist, such as This indicates a right-side delay; otherwise, the opposite applies.

[0156] (2) Calculate the torque difference:

[0157] The target torque difference is determined as Torque distribution is , ;

[0158] In the formula: This is the torque adjustment gain coefficient, primarily controlling the torque difference. and The proportional relationship affects the compensation response speed; , The torque distribution rule for the left and right motors after compensation is based on the right-side lag as an example: This indicates a decrease on the left side. Added to the right side.

[0159] (3) Dynamic amplitude limiting protection:

[0160] The application constraints are as follows ,in This represents the peak torque of the servo motor.

[0161] Furthermore, in this embodiment, the role of the sample set and its relation to... The relationships are as follows:

[0162] 1) Capable of supporting accurate identification of dynamic error parameters: The sample set stores N pairs of historical error data from different periods, which form the fitting angle. It provides continuous time-series data. Errors such as beam deformation and thermal drift in gantry machining centers exhibit time-varying characteristics. Single-period data is susceptible to random interference, while multi-period samples can capture error trends, enabling... The fitting results are more robust, thus accurately identifying the current linear offset and curvilinear deformation, avoiding the random bias of single-cycle data.

[0163] 2) Optimizable compensation instructions Output accuracy: Compensation command The core is to calculate the servo axis adjustment based on the current error state. The sample set provides historical evolution patterns of the error, allowing the system to predict error development trends, such as thermal deformation drift caused by continuous machining. It can not only compensate for current synchronization errors but also apply preventative adjustments in advance, enhancing the foresight of the compensation. Simultaneously, the sample set can be used to dynamically verify the compensation effect; by comparing the error convergence of historical samples, the compensation coefficients can be fine-tuned in real time, allowing for... The output is better adapted to the dynamic characteristics of the system, avoiding oscillations caused by over-adjustment.

[0164] 3) It can serve as a fundamental data source for constructing error models. The multi-condition error data accumulated in the sample set can be used to iteratively optimize the system dynamics model, making the model more closely match the actual operating characteristics of the machine tool, thereby enabling the generation of models based on the model. More accurate, achieving efficient suppression of positioning errors and asynchronous errors.

[0165] For example:

[0166] For the "data preprocessing" step in fitting the angle δ using historical data...

[0167] Real-time data acquisition: Data is acquired during each control cycle. , ,calculate and .

[0168] Historical data cache: stores data from the past N periods. Data pairs are used to form a sample set.

[0169] This sample set uses a fixed-length time-series circular cache structure, storing only the most recent N control cycles. Data pairs, such as N ranging from 5 to 20, are used to meet the real-time requirements of millisecond-level control cycles. Old data is automatically overwritten by data collected in the new cycle, ensuring that the samples always reflect the latest dynamic error characteristics of the system for subsequent angle fitting. Optimize compensation instructions.

[0170] The data schema fields are defined as shown in the table below:

[0171] Table 1, Data Schema Field Definitions:

[0172]

[0173] Table 2, Example of a parameter sequence table, assuming N=3 and the data are hypothetical, illustrates the correspondence between parameters in the sample set:

[0174]

[0175] When entering the k+2 cycle, the system will add new data such as (k+2, 0.0028, 0.0016) and automatically delete the oldest k-1 cycle data to maintain a fixed sample set length, providing continuous time-series support for dynamically identifying system deformation trends and accurately generating compensation instructions.

[0176] In this embodiment, the physical meaning of the parameter symbol table is as follows:

[0177] Table 3. Physical meaning of parameter symbols:

[0178]

[0179] In this embodiment:

[0180] 1) All gain parameters (Kp, Kτ, Kfb) need to be tuned through frequency sweep testing and step response analysis in the frequency domain;

[0181] 2) The compensation cycle must match the interpolation cycle of the CNC system to avoid control delay;

[0182] Furthermore, in this embodiment, Figure 5 The existing dual-drive gantry structure is shown, wherein dual-drive shaft 1 and dual-drive shaft 2 include servo motors, ball screws and guide rails, and the main shaft 4 is a sliding saddle type.

[0183] The core of the error compensation method for the dual-drive synchronous system of the gantry machining center of this invention is the construction of a synchronous control function module, which divides the dynamic characteristics of the dual-drive synchronous system into linear offset. and curvature deformation Two parameter values, and The angle between them is Given an objective function J, the goal is to minimize the threshold value. A dynamic model is established within the given range. Based on the objective function, a compensation algorithm is developed to approximate the threshold infinitely. This allows the dual-drive synchronization system to compensate for the positioning errors of the two synchronous shafts and the asynchrony errors of the two shafts, thereby improving synchronization accuracy.

[0184] in, d is the direct difference between the actual position feedback values ​​of the left and right servo axes. Figure 5 In the middle, it is intuitively reflected as This is the position deviation between the two drive axes directly obtained through the position feedback element. It belongs to the apparent synchronization error and is a direct reflection of the difference in the position tracking results of the two axes. This is the linear offset in the dynamic characteristics of a dual-drive synchronous system. It's a parameter describing the overall dynamic linear offset of the system, considering dynamic factors such as the linear deformation of the crossbeam 3 under load, acceleration, and thermal deformation, as well as the linear displacement caused by asymmetrical loads. It includes not only positional deviations but also the dynamic deformation of the system's mechanical structure, making it a core characteristic parameter for constructing the system's dynamic model. Therefore, compared to a horizontal straight line... Figure 3 In and It is based on an arbitrary curve variation. Figure 3 The dashed line indicates that this angle is also a dynamic angle whose size changes. .

[0185] Regarding the causes, d mainly originates from the positioning errors of the two axes (dual drive shaft 1 and dual drive shaft 2) and the real-time deviation of the synchronous control algorithm, which is a direct result of the difference in the position tracking accuracy of the axes. In addition to the influence of the positional deviation of the two axes, the formation of the linear displacement is also related to dynamic factors such as the mechanical structural deformation of the gantry machining center, the linear deformation of the crossbeam 3, the center of gravity shift caused by asymmetrical load, and the thermal deformation caused by temperature changes. It is a linear displacement manifestation under the combined action of multiple dynamic characteristics of the system.

[0186] Depending on the application, d is mainly used to monitor the synchronization error status of the two drive shafts in real time, serving as the direct basis for basic synchronization compensation. It is used to determine whether the current synchronization error exceeds the threshold and is a parameter at the real-time control level. It is one of the key parameters used to construct the system dynamics model and the objective function J, combined with the curvilinear deformation. Angle Parameters such as thresholds are set. It infinitely approximates this value to establish a precise compensation algorithm, thereby simultaneously compensating for the positioning error of the two axes and the asynchronous error of the dual-drive system. It is a core parameter for system-level error modeling and high-precision compensation, serving the construction of algorithms to improve overall synchronization accuracy.

[0187] Reference Figures 3-5 As shown, based on the standard coordinate system definition of a gantry machining center: the x-axis is the left-right movement direction of the dual-drive axis, the y-axis is the front-back movement direction of the main spindle, and the z-axis is the up-down feed direction of the main spindle.

[0188] Curved deformation Essentially, the bending deflection of the gantry beam 3 caused by uneven cutting forces and loads is primarily due to the bending along the z-axis, i.e., the vertical direction of the spindle 4. The load and cutting forces on the beam 3 supporting the spindle 4 mainly act along the z-direction, causing the beam 3 to bend in the z-x plane, resulting in a shift of the spindle 4 towards the z-position. The core error component.

[0189] The projection onto the xy plane refers to: The projection onto the xy-plane refers to projecting the spatial deformation of the bending beam 3 onto the xy-plane, corresponding to the differential displacement component along the x-axis. When the beam 3 bends, the left and right dual drive axes, x1 and x2, will experience relative x-direction positional deviations due to the beam deformation. This deviation is... The xy projection is used to quantify the asynchronous error in the x-direction of the dual-drive axis caused by the bending of the crossbeam 3, and is also used in the embodiment. offset from the line Forming an angle The core basis.

[0190] Projection in the z-direction: The principal component is the projection in the z-direction, which is the vertical deflection of the beam 3. It directly reflects the positioning offset of the spindle 4 in the z-direction and is a key error affecting the machining dimensional accuracy in the z-direction.

[0191] The x-axis differential displacement in the xy direction is the manifestation of the z-axis deflection on the dual drive shafts: when the beam bends in the z-direction, the left and right dual drive shafts will have a relative x-axis offset due to the beam deformation, i.e., the projection on the xy axis. This offset is the direct source of the dual drive synchronization error; while the z-axis deflection is the physical root cause of the beam deformation. In the above embodiment, the xy projection is identified... The parameters can be used to deduce the magnitude of the z-direction deflection, thereby simultaneously compensating for the x-direction synchronization error and the z-direction positioning error.

[0192] In the composite compensation instruction supplemented in this embodiment, the feedforward compensation is based on... , , Three parameters, among which yes and The included angle (including z-direction deflection and x-direction differential components) is essentially used to correlate the x and z bidirectional errors to achieve coordinated suppression of dual-drive synchronization error and z-direction positioning error, rather than ignoring z-direction deflection.

[0193] Unless otherwise specified, in this invention, terms such as "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe orientation or positional relationships in this invention are for illustrative purposes only and should not be construed as limiting this application. For those skilled in the art, the specific meaning of the above terms can be understood in conjunction with the accompanying drawings and according to the specific circumstances.

[0194] Unless otherwise explicitly specified and limited, the terms "set up," "connected," and "linked" in this invention should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0195] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications or equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for error compensation in a dual-drive synchronous system of a gantry machining center, characterized in that, include: The synchronization error of the dual-drive system of the gantry machining center is divided into two physical parameters: linear offset and curvilinear deformation. set up The linear offset represents the average positional deviation of the two drive shafts in the ideal direction of motion in the dual-drive system of the gantry machining center, reflecting the overall positioning inaccuracy problem; set up The amount of curvature deformation reflects the degree of angular torsion of the gantry beam of the gantry machining center caused by uneven stress or thermal deformation. Let δ be the angle between the linear offset and the curvilinear deformation, which represents the spatial orientation angle of the principal axis of deformation relative to the direction of motion; In the tangential coordinate system along the actual motion trajectory, and Projecting onto the x and y axes, we obtain the combined error components: Δx= + cosδ; Δy= sinδ; Where Δx is the total error in the x-direction; Δy is the lateral offset error in the y-direction; A target function J is set as a minimum distance measure between the current error state and the desired state of the double-drive system of the gantry machining center, and a reference benchmark C is defined 1x and C 1y , C 1x is an ideal compensation target value in the x direction, C 1y is an ideal compensation target value in the y direction, and an error square term is constructed: T=(Δx-C 1x ) 2 + (Δy-C 1y ) 2 ; make , representing the Euclidean distance from the current error vector to the target point; Introducing an ideal radius R, we construct an objective function in the form of a modulo operation: J = |R - t|; In the above formula, when t approaches R, J approaches 0. At this time, the dual-drive system of the gantry machining center is in the "tolerance error loop". If R is set as a minimum constant and ε is set as the threshold, when J < ε, it is considered that the system has entered the high-precision synchronization zone. By adjusting the controller output, J is continuously reduced, so that it can infinitely approach the threshold ε in the dynamic process, that is, J≤ε.

2. The error compensation method for a dual-drive synchronous system of a gantry machining center according to claim 1, characterized in that: The following steps are performed in each control cycle of the compensation algorithm: Step 1, Real-time Data Collection: Obtain the actual position feedback x1 and x2 of the left and right servo axes of the dual-drive system of the gantry machining center, and calculate the average position: x m = (x1 + x2) / 2, difference: d = x1 - x2; Step 2, Identification and : Assumption =x m - x ideal , where x ideal The positional error of the ideal trajectory; ∝ |d| / 2, where |d| / 2 represents the differential deformation. Fitting angle δ using historical data; Step 3, calculate Δx and Δy: The projection of the error onto the local coordinate system is obtained using trigonometric relationships; Step 4, update T, t, J Calculate the current objective function value and evaluate whether J≤ε is satisfied; Step 5, Generate compensation instructions If J > ε, then the speed and position commands of the two axes are adjusted through feedforward and feedback methods: right For those that are too large, apply translational correction; right The resulting Δy anomaly is addressed by implementing reverse torque ratio adjustment to suppress the torsional trend.

3. The error compensation method for a dual-drive synchronous system of a gantry machining center according to claim 2, characterized in that: The steps for fitting the angle δ using historical data are as follows: 1) Data preprocessing Real-time data acquisition: x1 and x2 are acquired in each control cycle, and calculations are performed. and ; Historical data cache: stores data from the past N periods. , Data pairs are used to form a sample set; 2) Model selection The estimated value of δ is fitted using linear regression or least squares method: =δ* +P, where P is the noise term, and the optimal δ is obtained by minimizing the sum of squared residuals: ； The mean of historical differential deformation: This reflects the long-term statistical characteristics of the deformation. This represents the mean of historical positioning errors: This reflects the long-term statistical characteristics of systematic errors.

Images