Horizontal machining center multi-station flexible clamping self-adaptive control method and system

Abstract

This invention discloses a multi-station flexible clamping adaptive control method and system for horizontal machining centers, comprising: establishing a digital model and acquiring the actual position and posture of the workpiece, calculating the position and posture deviation, and generating an initial lifting amount; extracting the wall thickness and stiffness distribution information of the workpiece at each station, combining the pre-estimated cutting force, calculating the deformation of the workpiece using finite element analysis, and solving the initial clamping force threshold of the clamping unit at each station with the minimum deformation as the objective function; during the machining process, acquiring the spindle current and power signals, the force signal of the force measuring support rod, and the acceleration signal of the worktable in real time, and performing time-frequency analysis to extract characteristic parameters; adjusting the clamping force according to the extracted characteristic parameters, executing a multi-station decoupling control algorithm to eliminate the mechanical coupling interference caused by the clamping force adjustment of adjacent stations; when changing surfaces for machining, pausing the clamping force and retaining the micro-force support, reacquiring the workpiece offset and performing position and posture compensation, and replanning the clamping force under the new machining surface.

Description

Technical Field

[0001] This invention relates to the field of machining control technology, specifically to a multi-station flexible clamping adaptive control method and system for horizontal machining centers. Background Technology

[0002] Horizontal machining centers, with their rotatable B-axis worktable, enable multi-faceted machining of workpieces and are widely used in the machining of complex box-shaped and shell-shaped parts. As multi-variety, small-batch production becomes mainstream, the demand for multi-station, flexible clamping is increasing. However, existing technologies have the following drawbacks in practical applications: Rigid clamping can cause workpiece deformation: Traditional fixtures often use fixed support and rigid clamping. When machining thin-walled or easily deformable parts, excessive clamping force can easily cause elastic or plastic deformation of the workpiece, which will spring back after machining and seriously affect the machining accuracy. Multi-station interference and resonance: When multiple workpieces are clamped at multiple stations simultaneously, the cutting forces at different stations may cause coupled vibration of the worktable. A single fixed clamping parameter cannot cope with the dynamic changes in the magnitude of the cutting forces at different stations, which can easily lead to a decrease in machining accuracy. Lack of adaptive adjustment capability: During the processing, as material is removed, the stiffness and stress distribution of the workpiece will change. Traditional clamping cannot sense and adjust the clamping force in real time, which leads to the workpiece deflection or vibration in the later stage of processing. Time-consuming recalibration during face-changing machining: After the horizontal machining center rotates along the B-axis, traditional flexible fixtures lack self-calibration capabilities, and the workpiece position shifts, requiring manual recalibration, which severely reduces the overall efficiency of multi-face machining. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the present invention aims to provide a multi-station flexible clamping adaptive control method and system for horizontal machining centers. Based on the digital model of the workpiece and real-time multi-source sensing signals, the system dynamically plans and independently adjusts the support force and clamping force of each station to achieve low-stress, high-precision, and decoupled flexible clamping, and automatically completes the pose compensation after B-axis face change.

[0004] To solve the above problems, the technical solution adopted by the present invention is as follows: An adaptive control method for multi-station flexible clamping in a horizontal machining center includes the following steps: S1. Initial pose acquisition and virtual pre-clamping of multi-station workpieces: acquire the CAD model and theoretical clamping position of the workpiece at each station, and establish a digital model of workpiece-fixture-machine table; acquire the actual pose of the workpiece by the pose sensor arranged on the worktable, calculate the pose deviation, and generate the initial lifting amount of the flexible support rod at each station. S2. Initial clamping force planning based on stiffness matching: Extract the wall thickness and stiffness distribution information of the workpiece at each station based on the CAD model of the workpiece, combine the estimated cutting force of the milling / boring process to be carried out, and use finite element analysis to calculate the deformation of the workpiece under different supports and clamping points; take the minimum deformation as the objective function to solve the initial clamping force threshold of the clamping unit at each station. S3. Real-time sensing and feature extraction of multi-source signals during machining: During machining, the spindle current / power signal, the force signal of the force support rod, and the acceleration signal of the worktable are collected in real time; time-frequency analysis is performed on the multi-source signals to extract characteristic parameters that characterize the sudden changes in cutting force, chatter, and local tool deflection of the workpiece at each station; S4. Adaptive adjustment of clamping force under multi-station coupling state: Based on the feature parameters extracted in step S3, determine the current processing state of each station and adjust the clamping force. Execute the multi-station decoupling control algorithm to eliminate the mechanical coupling interference caused by the clamping force adjustment of adjacent stations. S5, Adaptive Compensation of Position After B-Axis Rotation: When the horizontal machining center rotates the B-axis to perform face-changing machining, the clamping force is paused and a slight force support is maintained. The workpiece offset is re-acquired through the position sensor, and the offset is fed back to the CNC system for position compensation. Then, the system returns to step S2 to re-plan the clamping force under the new machining surface.

[0005] Preferably, when adjusting the clamping force, if the cutting force at a certain station increases and there is a risk of tool deflection, the clamping force and support force at that station are increased proportionally; if chatter is detected at a certain station, the clamping force at that station is reduced to release stress, and the stiffness damping of the support point is adjusted.

[0006] Preferably, generating the initial lifting amount includes: The actual three-dimensional coordinates of at least three non-collinear reference points on the workpiece surface are collected by a pose sensor and matched with the theoretical coordinates of the corresponding reference points in the digital model. The six-degree-of-freedom pose deviation of the workpiece relative to the theoretical clamping position is calculated, including translational deviations in the X / Y / Z directions and rotational deviations in the A / B / C axes. Based on the six-degree-of-freedom pose deviation, and taking the lifting range of each support rod and the contact constraint between the rod and the bottom surface of the workpiece as boundary conditions, an inverse kinematics solution model for workpiece pose correction is established. The initial lifting amount of each flexible support rod is obtained by solving, so that the positional deviation of the workpiece after correction is less than the preset clamping accuracy threshold, and all support rods maintain surface contact with the bottom surface of the workpiece without gaps. In the inverse kinematics solution model, a worktable flatness error compensation parameter is introduced to pre-substitute the worktable's own flatness deviation into the initial lifting amount calculation, so as to offset the systematic error caused by worktable processing or wear. The clamping accuracy threshold is dynamically set according to the accuracy requirements of the corresponding workstation's machining process: the threshold for rough machining is set to 0.05mm-0.1mm, and the threshold for finish machining is set to 0.005mm-0.02mm.

[0007] Preferably, the calculation of deformation includes: Import the CAD model of the workpiece, and mesh the model according to the actual working conditions. Refine the mesh for thin-walled areas, clamping contact areas, and cutting areas. The mesh size is 1 / 3 to 1 / 5 of that for non-critical areas. In the finite element model, a fixed displacement constraint is applied to the position of the support rod, a normal pressure load is applied to the position of the clamping unit, and a three-dimensional dynamic cutting force load is applied to the cutting position of the machining process. Static simulation is performed to extract the maximum normal deformation of the workpiece's machined surface, the stress distribution in the clamping contact area, and the reaction force data of the support points as the deformation calculation results.

[0008] Preferably, solving for the initial clamping force threshold includes: A multi-objective optimization function is constructed, with the minimum maximum normal deformation of the workpiece machining surface as the core objective, and the minimum total clamping force and the minimum stress difference between clamping points as auxiliary optimization objectives. The objective weights are set to 0.6, 0.2, and 0.2, respectively. Set constraints: including the clamping force adjustment range constraint of each clamping unit, the stress constraint that the maximum stress in the workpiece contact area does not exceed 70% of the material yield strength, and the full contact constraint that the reaction force of all support points is greater than 0. The objective function is solved iteratively using a particle swarm optimization algorithm. The iteration terminates when the change in deformation is less than 1 μm after three consecutive iterations. The optimal initial clamping force threshold for each clamping unit is then output.

[0009] Preferably, when executing the multi-station decoupling control algorithm, it includes: Beforehand, mechanical coupling calibration of the workbench is performed. Under no-load conditions, a simulated clamping force with gradient changes is applied to each station one by one. The force change values ​​of the support points of all other stations are collected to construct a global coupling coefficient matrix, with matrix element K. ij This represents the coupling interference force experienced by the j-th workstation when a unit clamping force is applied to the i-th workstation; During real-time control, when a workstation needs to adjust the clamping force, the estimated value of the coupling interference of the clamping force adjustment on all other adjacent workstations is calculated based on the global coupling coefficient matrix. The compensation adjustment is synchronously output to the support rods of all other adjacent workstations to counteract the influence of coupling interference force and ensure that the support force fluctuation of all non-adjustment workstations is less than the preset threshold.

[0010] Preferably, the digital model established in step S1 includes workpiece material properties, fixture support point and clamping point position parameters, and worktable mechanical coupling coefficient.

[0011] Preferably, the feature parameters extracted in step S3 include: the time-domain amplitude of the cutting force, the chatter characteristic frequency of the vibration signal in the frequency domain, the spindle power change rate, and the dynamic fluctuation of the force on the force-measuring support rod.

[0012] Preferably, the position compensation after B-axis rotation in step S5 only applies to the overall offset of the workpiece, without adjusting the position of the support rod. After the compensation is completed, the clamping force is recalculated based on the cutting force parameters of the new machining surface.

[0013] A multi-station flexible clamping adaptive control system for a horizontal machining center includes: The pose sensors are distributed at each station of the worktable to obtain the six-degree-of-freedom pose deviation of the workpiece at the initial stage and after the B-axis rotation. The multi-axis flexible clamping execution module includes arrayed servo-electric support rods and servo clamping units, which can independently control the position and force of each support point and clamping point; A multi-source signal acquisition module is integrated into the spindle, fixture, and worktable for real-time acquisition of current, force, and vibration signals. The adaptive control terminal, which incorporates a stiffness matching algorithm and a multi-station decoupling control algorithm, communicates with the posture sensor, the multi-axis flexible fixture execution module, the multi-source signal acquisition module, and the machine tool CNC system to execute the aforementioned adaptive control methods.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: Deformation prevention and low-stress clamping: By planning the initial clamping force based on stiffness matching, the optimal clamping force threshold is solved with the goal of minimizing workpiece deformation, and the clamping force is adaptively adjusted according to the cutting state during machining, which avoids the deformation of thin-walled parts caused by traditional rigid clamping and significantly improves machining accuracy. Independent and decoupled control of multiple stations: For multi-station machining scenarios of horizontal machining centers, a global coupling coefficient matrix is ​​constructed by pre-calibration and a decoupled control algorithm is executed to realize independent adaptive adjustment of clamping force of each station, effectively eliminating mechanical coupling interference caused by clamping force adjustment between multiple stations and avoiding resonance and workpiece displacement; B-axis rotation fast adaptive: Utilizing pose perception and CNC coordinate dynamic compensation mechanism, after B-axis rotation, only micro-force support needs to be retained to quickly remeasure and compensate for the overall offset. There is no need for manual secondary dialing and alignment, nor is there a need to adjust the position of the support rod, which greatly shortens the surface change auxiliary time and improves the efficiency of multi-face processing. Suppressing chatter and tool deflection: By acquiring multi-source signals in real time and extracting characteristic parameters, it can accurately identify sudden changes in cutting force and chatter risks. By increasing the clamping force proportionally, tool deflection can be suppressed, or the clamping force can be reduced and the stiffness and damping of the support point can be adjusted to release stress and suppress chatter, effectively solving the problems of tool vibration and tool deflection in the later stages of machining. Highly adaptable to flexible production: A closed-loop control logic of "virtual pre-clamping - initial stiffness matching - processing status perception - dynamic decoupling adjustment - automatic surface change compensation" has been constructed, realizing the dynamic adaptation of clamping parameters and processing process, and adapting to the needs of flexible production scenarios with multiple varieties and small batches.

[0015] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0016] Figure 1 This is a logic block diagram of the multi-station flexible clamping adaptive control method for horizontal machining centers according to an embodiment of the present invention. Figure 2 This is a logic block diagram for calculating the initial lifting amount in an embodiment of the present invention; Figure 3 This is a block diagram of the finite element deformation calculation logic in an embodiment of the present invention. Figure 4 This is a logic block diagram for solving the initial clamping force threshold according to an embodiment of the present invention; Figure 5 This is a logic block diagram of the multi-station decoupling control algorithm according to an embodiment of the present invention; Figure 6 This is an interactive diagram of the multi-station flexible clamping adaptive control system module for a horizontal machining center according to an embodiment of the present invention. Detailed Implementation

[0017] Example 1, see Figure 1 This invention provides a multi-station flexible clamping adaptive control method for horizontal machining centers, comprising the following steps: S1. Initial pose acquisition and virtual pre-clamping of multi-station workpieces: acquire the CAD model and theoretical clamping position of the workpiece at each station, and establish a digital model of workpiece-fixture-machine table; acquire the actual pose of the workpiece by the pose sensor arranged on the worktable, calculate the pose deviation, and generate the initial lifting amount of the flexible support rod at each station. S2. Initial clamping force planning based on stiffness matching: Extract the wall thickness and stiffness distribution information of the workpiece at each station based on the CAD model of the workpiece, combine the estimated cutting force of the milling / boring process to be carried out, and use finite element analysis to calculate the deformation of the workpiece under different supports and clamping points; take the minimum deformation as the objective function to solve the initial clamping force threshold of the clamping unit at each station. S3. Real-time sensing and feature extraction of multi-source signals during machining: During machining, the spindle current / power signal, the force signal of the force support rod, and the acceleration signal of the worktable are collected in real time; time-frequency analysis is performed on the multi-source signals to extract characteristic parameters that characterize the sudden changes in cutting force, chatter, and local tool deflection of the workpiece at each station; S4. Adaptive adjustment of clamping force under multi-station coupling state: Based on the feature parameters extracted in step S3, determine the current processing state of each station and adjust the clamping force. Execute the multi-station decoupling control algorithm to eliminate the mechanical coupling interference caused by the clamping force adjustment of adjacent stations. S5, Adaptive Compensation of Position After B-Axis Rotation: When the horizontal machining center rotates the B-axis to perform face-changing machining, the clamping force is paused and a slight force support is maintained. The workpiece offset is re-acquired through the position sensor, and the offset is fed back to the CNC system for position compensation. Then, the system returns to step S2 to re-plan the clamping force under the new machining surface.

[0018] Specifically, this method addresses technical issues such as deformation of thin-walled parts, multi-station coupled vibration, inability to dynamically adapt clamping parameters, and low efficiency of B-axis face-changing alignment in multi-station machining scenarios of horizontal machining centers. It constructs a closed-loop control logic of "virtual pre-clamping - initial stiffness matching - machining state perception - dynamic decoupling adjustment - automatic face-changing compensation" to achieve dynamic adaptation of clamping parameters and the machining process. The principle is as follows: (1) Initial pose acquisition and virtual pre-clamping of multiple workstations: rapid pose correction based on digital model The core of this step is to achieve a precise mapping between the actual clamping state of the workpiece and the theoretical model, replacing the traditional manual dial indicator alignment: First, integrate the workpiece CAD model, fixture layout, and worktable physical parameters to construct a digital model of "workpiece-fixture-worktable" that maps the geometric dimensions, material properties, and mechanical coupling characteristics of the physical scene. Using non-contact laser pose sensors arranged on the worktable, the actual three-dimensional coordinates of at least three non-collinear reference points on the workpiece surface are collected and matched with the theoretical coordinates of the CAD model to calculate the pose deviation of the workpiece in six degrees of freedom: X / Y / Z translation and A / B / C rotation. With the constraints of the support rod stroke range and full contact between the rod and the workpiece bottom surface, an inverse kinematics solution model is established. At the same time, the worktable flatness error compensation parameter is introduced to automatically calculate the initial lifting amount of each flexible support rod, and control the workpiece position deviation within the accuracy threshold corresponding to the process (0.05~0.1mm for roughing and 0.005~0.02mm for finishing), so as to achieve gapless pre-clamping.

[0019] (2) Initial clamping force planning based on stiffness matching: pre-optimization of clamping parameters with the goal of minimizing deformation The core of this step is to resolve the contradiction in rigid clamping: "excessive clamping force leads to workpiece deformation, while insufficient clamping force leads to machining loosening." First, extract the wall thickness distribution data from the workpiece CAD model, calculate the stiffness characteristics at different locations of the workpiece, and identify thin-walled, easily deformable weak areas. Based on the estimated cutting force parameters of the current process (the three-dimensional cutting force range of milling / boring), a finite element simulation model of the workpiece is established. The mesh of the thin-walled area, clamping contact area, and machining area is refined to simulate the workpiece deformation and stress distribution under different combinations of support force and clamping force. A multi-objective optimization function is constructed: the core objective is to minimize the maximum normal deformation of the machined surface (weight 0.6), and the auxiliary objectives are to minimize the total clamping force and ensure uniform stress at each clamping point (weight 0.2 each). At the same time, the constraints of "clamping force within the adjustment range of the fixture, workpiece contact stress not exceeding 70% of the material yield strength, and all support points being under positive force" are satisfied. The optimal initial clamping force threshold of each station clamping unit is obtained by iterative solution through particle swarm optimization algorithm, so as to achieve precise matching between clamping stiffness, workpiece stiffness, and cutting force.

[0020] (3) Real-time sensing and feature extraction of multi-source signals during processing: rapid identification of processing anomalies The core of this step is to build a multi-dimensional processing status perception system to achieve early warning of abnormal states: Three types of signals are acquired simultaneously: spindle current / power signal (reflecting changes in cutting load), force signal from the force sensor inside the support rod (reflecting workpiece deformation under force), and vibration signal from the table acceleration sensor (reflecting chatter characteristics). The acquired time-domain signal is converted to the frequency domain, and four types of feature parameters are extracted: cutting force time-domain amplitude (to determine load abrupt changes), chatter characteristic frequency in the frequency domain of vibration signal (to identify chatter risk), spindle power change rate (to determine tool deflection trend), and dynamic fluctuation of support force (to determine workpiece loosening / deformation), so as to achieve quantitative characterization of three typical machining anomalies: cutting force abrupt changes, chatter, and local tool deflection.

[0021] (4) Adaptive adjustment of clamping force under multi-station coupling: decoupled dynamic closed-loop control The core of this step is to resolve the mutual interference problem in clamping force adjustments during multi-station machining, and to achieve dynamic adaptation of clamping parameters: Pre-calibration stage: Under no-load conditions, gradient clamping force is applied to each station one by one, the force changes of the support points of the remaining stations are collected, a global coupling coefficient matrix is ​​constructed, and the interference of the clamping force change of the i-th station on the j-th station is quantified. Real-time adjustment phase: Based on the extracted feature parameters, the processing status is determined, and a differentiated adjustment strategy is adopted. If an increase in cutting force or a higher risk of tool deflection is detected: proportionally increase the clamping force and support force at that station to suppress workpiece deformation during machining; If chatter is detected: appropriately reduce the clamping force at this station to release internal stress, and at the same time adjust the damping parameters of the support rod to suppress vibration propagation; Decoupling compensation: When the clamping force of a certain station is adjusted, the estimated interference of the adjustment to all other stations is calculated based on the coupling coefficient matrix. The compensation adjustment amount is output to the support rods of the other stations simultaneously to ensure that the support force fluctuation of the non-adjustment station is less than the preset threshold, and to avoid resonance or workpiece displacement caused by mechanical coupling between multiple stations.

[0022] (5) Adaptive position compensation after B-axis rotation: No manual alignment is required for face-changing machining. The core of this step is to solve the problem of automatic compensation for workpiece offset after the B-axis rotation of a horizontal machining center, thereby improving the efficiency of multi-faceted machining: After the B-axis rotation completes the face change, the system pauses the clamping force and retains a slight support to prevent workpiece displacement. The position sensor completes the workpiece position remeasurement within 1 second, calculates the overall workpiece offset, and directly feeds it back to the machine tool CNC system for automatic compensation of the machining coordinate system, without the need for manual realignment or adjustment of the support rod position. For the new machining surface, the stiffness matching algorithm in step S2 is called again, and the corresponding clamping force is planned based on the cutting force parameters of the new process. The adaptive control process of the next machining surface is directly entered, realizing the seamless connection of multi-face machining.

[0023] Compared with traditional rigid clamping, this method can significantly reduce machining deformation of thin-walled parts, improve the machining efficiency of multi-stations, significantly reduce B-axis face-changing and alignment time, and effectively suppress machining chatter and tool deflection problems, making it suitable for flexible production scenarios with multiple varieties and small batches.

[0024] In one possible embodiment, when adjusting the clamping force, if the cutting force at a certain station increases and there is a risk of tool deflection, the clamping force and support force at that station are increased proportionally; if chatter is detected at a certain station, the clamping force at that station is reduced to release stress, and the stiffness damping of the support point is adjusted.

[0025] Specifically, the system detects a sudden increase in the time-domain amplitude of the cutting force and an abnormal increase in the rate of change of the spindle power by acquiring multi-source signals. Combined with the workpiece stiffness distribution data, it determines that the current cutting load exceeds the bearing threshold of the initial clamping force, and there is a risk that the workpiece will be lifted by the tool and the machining surface size will be too small.

[0026] Adjustment principle: The strategy of "proportional matching and enhancement of clamping force and support force" is adopted: the clamping force is used to enhance the friction between the workpiece and the fixture contact surface to prevent the workpiece from sliding horizontally under the action of cutting force; the support force is simultaneously enhanced to strengthen the local stiffness of the workpiece machining area and offset the elastic deformation of the workpiece caused by the cutting normal force. The ratio of the two is dynamically determined according to the pre-calculated workpiece stiffness distribution parameters to ensure that the workpiece is subjected to balanced force.

[0027] During the adjustment process, the stress in the contact area of ​​the workpiece is checked in real time to ensure that the total stress does not exceed 70% of the material's yield strength, thus avoiding excessive clamping force that could lead to localized plastic deformation of the workpiece.

[0028] The system performs time-frequency analysis on the table acceleration signal and detects chatter characteristic frequencies (coupled frequencies of cutting harmonics and system natural frequencies) in the frequency domain with amplitudes exceeding the warning threshold. It determines that the tool-workpiece-fixture system is experiencing unstable self-excited vibration, which can easily lead to surface chatter marks and tool breakage.

[0029] Adjustment principle: The "stress release + damping matching" combination strategy is adopted: appropriately reducing the clamping force is to release the internal stress accumulated in the workpiece during clamping and early cutting process, and break the energy feedback cycle of chatter; the damping parameters of the support rod are adjusted simultaneously to increase the dissipation of vibration energy, block the propagation path of vibration in the workpiece-fixture system, and quickly suppress the chatter amplitude.

[0030] This adjustment differs from the traditional logic of "increasing clamping force to suppress vibration." It is primarily aimed at the coupled flutter scenario caused by excessive clamping stress in thin-walled parts, and achieves vibration source control by actively reducing the load.

[0031] See Figure 2 In one possible embodiment, the specific steps for calculating the pose deviation and generating the initial lifting amount of the flexible support top rod at each station include: The actual three-dimensional coordinates of at least three non-collinear reference points on the workpiece surface are collected by a pose sensor and matched with the theoretical coordinates of the corresponding reference points in the digital model. The six-degree-of-freedom pose deviation of the workpiece relative to the theoretical clamping position is calculated, including translational deviations in the X / Y / Z directions and rotational deviations in the A / B / C axes. Based on the six-degree-of-freedom pose deviation, and taking the lifting range of each support rod and the contact constraint between the rod and the bottom surface of the workpiece as boundary conditions, an inverse kinematics solution model for workpiece pose correction is established. The initial lifting amount of each flexible support rod is obtained by solving, so that the positional deviation of the workpiece after correction is less than the preset clamping accuracy threshold, and all support rods maintain surface contact with the bottom surface of the workpiece without gap.

[0032] Specifically, the core principle of this step is automatic workpiece pose correction based on geometric matching and kinematic solution. It is a core preliminary step in the flexible clamping process of horizontal machining centers. Its purpose is to replace traditional manual dial indicator alignment and achieve high-precision, gapless adaptation between the actual workpiece position and the theoretical clamping position. The specific principle is as follows: (1) Calculation of six-degree-of-freedom pose deviation: accurate mapping between actual pose and theoretical model The essence of this step is to solve for the pose matching of rigid bodies in three-dimensional space: The pose of a rigid body in three-dimensional space is uniquely determined by six independent parameters: translational deviations in the X, Y, and Z directions, and rotational deviations around the X-axis (A-axis), Y-axis (B-axis), and Z-axis (C-axis). The spatial pose of a rigid body can be uniquely determined by the coordinates of at least three non-collinear reference points: by collecting the actual coordinates of predefined reference points on the workpiece by a pose sensor, and performing point set matching with the theoretical coordinates of the corresponding reference points in the digital model (usually using the singular value decomposition (SVD) algorithm to solve for the optimal matching matrix), all six pose deviations of the workpiece's current actual position relative to the theoretical clamping position can be calculated.

[0033] This process can be understood as: "anchoring" the actual spatial state of the workpiece through three or more reference points, and quantifying its geometric deviation from the ideal clamping state.

[0034] (2) Construction of inverse kinematics solution model: conversion of position deviation to push rod adjustment amount This step involves reversing the kinematics of the actuator input from the target pose: The workpiece is supported by multiple flexible support rods. The change in the lifting amount of each rod will cause the workpiece to produce a corresponding change in posture. The two satisfy a certain kinematic mapping relationship. The model's core objective is to "eliminate the six-degree-of-freedom pose deviations obtained in the first step of calculation," while simultaneously setting two types of boundary constraints: Physical stroke constraint: The lifting amount of each push rod must be within the upper and lower limits of its servo adjustment to avoid overtravel; Contact constraint: All push rods must maintain surface contact with the bottom surface of the workpiece without gaps to ensure support rigidity and avoid machining deformation caused by local suspension.

[0035] To address the flatness wear error caused by long-term use of the workbench, the model can also pre-incorporate flatness compensation parameters to include system errors in the calculation, thereby further improving the correction accuracy.

[0036] (3) Initial lifting amount solution: Output the optimal solution that satisfies the accuracy and constraints. Solving the constructed inverse kinematics model, the final output of the initial lifting amount of each push rod must simultaneously satisfy two judgment conditions: Accuracy requirements: The positional deviation of the workpiece after lifting and correction is less than the clamping accuracy threshold corresponding to the process (0.05~0.1mm for roughing and 0.005~0.02mm for finishing), which meets the alignment requirements for subsequent processing. Mechanical conditions: All support rods maintain stable contact with the bottom surface of the workpiece, with no local gaps or over-the-top contact, providing a uniform support base for subsequent clamping force loading.

[0037] The calculated lifting amount is directly sent to each servo push rod for execution, which can realize the automatic alignment of the workpiece without manual intervention.

[0038] The core advantage of this step is that it transforms the traditional alignment process, which relies on human experience, into a quantifiable and automated geometric solution process. This adapts to the rapid clamping requirements of different workpieces in multi-variety, small-batch production, and the alignment accuracy and efficiency are significantly higher than those of manual operation.

[0039] In one possible embodiment, the inverse kinematics solution model simultaneously introduces a worktable flatness error compensation parameter, which substitutes the flatness deviation of the worktable itself into the calculation of the initial lifting amount in advance, so as to offset the systematic error caused by worktable processing or wear. The clamping accuracy threshold is dynamically set according to the machining process accuracy requirements of the corresponding workstation: the threshold for rough machining is set to 0.05mm-0.1mm, and the threshold for finish machining is set to 0.005mm-0.02mm.

[0040] Specifically, this step is the core optimization mechanism of the initial pose correction stage of flexible clamping in horizontal machining centers. By combining pre-compensation for system errors with dynamic adaptation of accuracy thresholds, the workpiece clamping accuracy is matched with the machining requirements.

[0041] (1) Workbench flatness error compensation mechanism This mechanism addresses the pre-emptive elimination of inherent systematic errors in machine tool worktables. Its core logic involves incorporating static systematic errors into the boundary conditions of the kinematic solution to prevent clamping deviations caused by error accumulation. The specific process is as follows: Sources of error: During long-term use, the machine tool worktable may develop flatness deviations due to factors such as residual stress from machining and assembly, long-term load of heavy objects, and cutting wear. These deviations are fixed or slowly varying systematic errors. If they are not compensated, they will be directly added to the initial positional deviation of the workpiece, resulting in a decrease in the final clamping accuracy.

[0042] Parameter pre-calibration: The flatness of the entire worktable surface is scanned in advance using a coordinate measuring machine or a laser flatness tester to generate height deviation grid data of the worktable surface. Each grid point corresponds to the deviation value between the actual height and the theoretical plane at a specific position on the worktable, forming a worktable flatness error compensation parameter library.

[0043] Error embedding solution: When constructing the inverse kinematics solution model, the flatness deviation value of the worktable under the corresponding work station is taken as a known quantity and substituted into the calculation process of the initial lifting amount of the support rod in advance. That is, the target lifting amount of the rod not only needs to offset the workpiece's own posture deviation, but also needs to compensate for the height deviation of the worktable at the support point. In the end, the parallelism error between the bottom surface of the workpiece and the theoretical clamping plane is offset in advance, avoiding the transmission of systematic errors caused by worktable wear to the workpiece.

[0044] (2) Dynamic setting mechanism for clamping accuracy threshold This mechanism follows the principle of balancing accuracy and efficiency, and sets the allowable range of clamping errors differently according to the actual needs of the machining process, so as to avoid the loss of clamping efficiency caused by excessive accuracy requirements. Threshold adaptation logic: The clamping accuracy threshold is the maximum allowable deviation after workpiece pose correction, and its setting is directly linked to the accuracy requirements of subsequent machining processes. Rough machining process: The goal of this stage is to quickly remove excess material. The machining allowance is usually above 1mm, which allows for a relatively large machining error. Therefore, the clamping accuracy threshold is set to 0.05mm-0.1mm, which can meet the positioning requirements of rough machining, reduce the complexity of solving the initial lifting amount, and improve the clamping efficiency of multi-station.

[0045] Finishing process: This stage requires ensuring the final dimensional accuracy and geometric tolerances of the parts. The machining allowance is usually within 0.1mm, and the positioning accuracy requirements are extremely high. Therefore, the clamping accuracy threshold is tightened to 0.005mm-0.02mm, and the initial position deviation is strictly controlled to avoid clamping errors occupying the finishing tolerance allowance and to ensure the final machining accuracy.

[0046] Dynamic scheduling is achieved by reading the process type identifier in the machining process file and automatically matching the clamping accuracy threshold of the corresponding station as the termination condition for solving the inverse kinematics model. When the lifting amount obtained by the solution can make the workpiece pose deviation less than the corresponding threshold, it can be judged as a qualified solution, thus realizing the automatic adaptation of clamping accuracy and machining requirements.

[0047] The two mechanisms work together to eliminate the static system error of the machine tool itself and avoid unnecessary precision redundancy, thereby improving the adaptability and reliability of multi-station flexible clamping.

[0048] See Figure 3 In one possible embodiment, the specific steps for calculating the deformation of the workpiece under different support and clamping points using finite element analysis include: Import the CAD model of the workpiece, and mesh the model according to the actual working conditions. Refine the mesh for thin-walled areas, clamping contact areas, and cutting areas. The mesh size is 1 / 3 to 1 / 5 of that for non-critical areas. In the finite element model, a fixed displacement constraint is applied to the position of the support rod, a normal pressure load is applied to the position of the clamping unit, and a three-dimensional dynamic cutting force load is applied to the cutting position of the machining process. Static simulation is performed to extract the maximum normal deformation of the workpiece's machined surface, the stress distribution in the clamping contact area, and the reaction force data of the support points as the deformation calculation results.

[0049] Specifically, this step uses numerical simulation to quantify the mechanical response of the workpiece under the combined action of clamping and cutting loads, providing a quantitative basis for subsequent clamping force optimization.

[0050] (1) Construction of high-precision finite element model: differentiated mesh generation that balances computational accuracy and efficiency The core of this step is to transform the three-dimensional geometric model of the workpiece into a discretized numerical model capable of mechanical calculations, thus solving the balance between "overall computational efficiency" and "accuracy in key areas": Model import and preprocessing: Import the 3D CAD model of the workpiece and perform geometric cleanup, deleting details such as fillets and small holes that have minimal impact on mechanical calculations to avoid unnecessary computation; at the same time, embed the workpiece material properties (elastic modulus, Poisson's ratio, yield strength, etc.) to ensure that the simulation parameters are consistent with the actual workpiece.

[0051] Differentiated mesh generation logic: A non-uniform mesh generation strategy is adopted to refine the mesh in three types of mechanically sensitive areas, with the mesh size set to 1 / 3-1 / 5 of that in non-critical areas. Thin-walled regions have low stiffness and are the main areas where processing deformation occurs. Dense meshing can accurately capture local elastic deformation. Stress concentration exists in the clamping contact area; a denser mesh can accurately simulate the transmission path of clamping force and stress distribution. The cutting area directly bears the cutting load, and a denser mesh can ensure the accuracy of deformation calculations on the machined surface.

[0052] Non-critical areas (such as rigid thick walls and non-processed areas) use larger mesh sizes, which significantly reduces the overall simulation computation load without affecting the overall calculation accuracy, and meets the needs of rapid iterative calculations at multiple workstations.

[0053] (2) Boundary conditions and load application: 1:1 digital mapping of physical clamping and machining scenarios The core of this step is to reproduce the mechanical boundaries during the actual machining process in the finite element model, ensuring the consistency between the simulation results and the real working conditions. Constraint application: Apply fixed displacement constraints (restricting the degrees of freedom in the X / Y / Z directions) at the position of the corresponding support rod to simulate the rigid support effect of the support rod on the workpiece and reproduce the boundary conditions of the workpiece being supported and positioned in actual clamping.

[0054] Clamping load application: Apply a normal pressure load at the position of the corresponding clamping unit. The load magnitude covers the entire adjustment range of the clamping force of the fixture, simulating the stress state of the workpiece under different clamping forces.

[0055] Cutting load application: Apply a three-dimensional dynamic cutting force load (cutting force components in the X, Y, and Z directions) to the cutting position of the current process. The load parameters are obtained from the actual cutting force database or empirical formula of the corresponding milling / boring process to reproduce the dynamic force of the tool on the workpiece during the machining process.

[0056] (3) Static solution and feature extraction: Output of key mechanical indicators for clamping optimization The core of this step is to obtain the complete mechanical response of the workpiece under a set load by solving the elasticity equilibrium equation, and to extract the key indicators directly related to clamping optimization: Solution logic: Based on the linear statics theory under the assumption of small deformation, solve the discretized finite element equilibrium equations (load vector = overall stiffness matrix × nodal displacement vector) to obtain the physical quantities such as displacement and stress of all nodes.

[0057] Key Indicator Extraction: Three types of calculation results directly serving clamping force optimization were extracted: Maximum normal deformation of the machined surface: directly reflects the dimensional deviation of the workpiece after machining, and is the core objective of subsequent optimization; Stress distribution value in the clamping contact area: used to determine whether the clamping force is too large, causing local yielding or indentation damage to the workpiece; The reaction force data of the support point is used to verify whether the support rod is in full contact with the workpiece and to avoid support failure caused by partial suspension.

[0058] The output results are directly used as input parameters for subsequent multi-objective optimization functions, providing a quantitative basis for determining the optimal clamping force.

[0059] This finite element calculation process transforms the traditional manual experience-based selection of clamping parameters into a quantifiable and reproducible numerical analysis process. It is the core foundation for achieving "precise matching of clamping stiffness with workpiece stiffness and cutting force" and can significantly improve the deformation control accuracy of initial clamping force planning.

[0060] See Figure 4 In one possible embodiment, the specific steps for solving the initial clamping force threshold of each station clamping unit, with the minimum deformation as the objective function, include: A multi-objective optimization function is constructed, with the minimum maximum normal deformation of the workpiece machining surface as the core objective, and the minimum total clamping force and the minimum stress difference between clamping points as auxiliary optimization objectives. The objective weights are set to 0.6, 0.2, and 0.2, respectively. Set constraints: including the clamping force adjustment range constraint of each clamping unit, the stress constraint that the maximum stress in the workpiece contact area does not exceed 70% of the material yield strength, and the full contact constraint that the reaction force of all support points is greater than 0. The objective function is solved iteratively using a particle swarm optimization algorithm. The iteration terminates when the change in deformation is less than 1 μm after three consecutive iterations. The optimal initial clamping force threshold for each clamping unit is then output.

[0061] Specifically, this step is the core step of pre-optimizing the initial clamping parameters based on stiffness matching, which aims to solve the technical problems in traditional rigid clamping where "excessive clamping force leads to workpiece deformation and insufficient clamping force leads to loosening during processing".

[0062] (1) Construction of multi-objective optimization function: Quantitative evaluation system of clamping effect The core of this step is to transform the clamping requirements of "low deformation, low stress, and uniform stress" into calculable mathematical objectives, and to achieve priority balance among multiple objectives through weight allocation: Core objective (weight 0.6: Minimize the maximum normal deformation of the workpiece's machined surface): The normal deformation of the machined surface is a core indicator that directly determines the final dimensional accuracy and geometric tolerance of the part. It is given the highest priority as an objective to ensure that the machining error is minimized, which meets the core requirements of precision machining.

[0063] Auxiliary objective 1 (weight 0.2: minimum total clamping force): While meeting deformation control requirements, minimize the overall clamping force to avoid excessive clamping stress causing residual deformation of the workpiece, while reducing the load wear of the fixture and extending the service life of the fixture.

[0064] Auxiliary objective 2 (weight 0.2: minimum stress difference between clamping points): ensure that the clamping force is evenly distributed among multiple clamping points, avoid local stress concentration that could cause workpiece damage or plastic deformation, and make the workpiece more evenly stressed, thus reducing the risk of off-center loading during processing.

[0065] The three factors are weighted and summed to form a unified objective function, which not only ensures the core requirement of machining accuracy, but also takes into account the process requirements of clamping stress control and force uniformity.

[0066] (2) Constraint setting: Boundary guarantee of clamping feasibility The purpose of this step is to limit the feasible region of the optimization solution, ensuring that the clamping force parameters obtained from the solution are both within the capability range of the physical equipment and meet the process safety requirements of workpiece machining. Clamping force adjustment range constraint: The clamping force of each clamping unit must be within the rated output range of the servo system to ensure that the output parameters can be actually executed and to avoid invalid solutions that exceed the range.

[0067] Stress constraint (maximum stress in the contact area ≤ 70% of the material yield strength): The upper limit of clamping force is limited from the perspective of material mechanical properties. The 70% safety factor not only avoids plastic deformation and surface indentation damage in the workpiece contact area caused by excessive clamping force, but also reserves the stress margin for the superposition of cutting force during the machining process, ensuring that the clamping process is always within the elastic deformation range.

[0068] Full contact constraint (reaction force at all support points > 0): ensures that the workpiece always maintains surface contact with all support rods under clamping force, without local suspension, avoids support failure leading to decreased workpiece stiffness, vibration or displacement during processing, and ensures the overall rigidity of the clamping system.

[0069] All constraints together constitute the feasible space of the solution, ensuring that the optimization results are both reasonable and feasible.

[0070] (3) Particle swarm optimization algorithm for iterative solution: efficient search for the global optimum This step employs a heuristic intelligent algorithm to quickly search for the optimal solution of the objective function within the feasible region, adapting to multi-parameter, nonlinear clamping optimization scenarios: Algorithm adaptability: Particle Swarm Optimization (PSO) is based on the iterative search logic of swarm intelligence. It does not rely on the gradient information of the objective function and has good global search capabilities for multivariable and nonlinear engineering problems such as clamping optimization. It can avoid getting trapped in local optima and its solution efficiency is much higher than that of traditional enumeration methods.

[0071] The termination condition setting, which is a convergence criterion that the deformation change value of three consecutive iterations is less than 1μm, ensures that the solution accuracy meets the deformation control requirements of precision machining (the deformation fluctuation of 1μm is much smaller than the tolerance requirements of conventional precision machining), while avoiding unnecessary over-iteration, thus balancing computational efficiency and accuracy.

[0072] Output results: The final output of the initial clamping force threshold of each clamping unit is the optimal combination of clamping parameters under the current workpiece stiffness and cutting load conditions. It achieves a multi-objective balance of "minimum deformation, lowest stress, and uniform force", providing the optimal initial clamping state for subsequent machining processes.

[0073] This solution process transforms the traditional manual experience-based selection of clamping parameters into a quantifiable and reproducible numerical optimization process. It is the core technical support for achieving "precise matching of clamping stiffness with workpiece stiffness and cutting force", which can significantly reduce clamping deformation of thin-walled parts and improve the consistency of machining accuracy.

[0074] See Figure 5 In one possible embodiment, the specific steps of executing the multi-station decoupling control algorithm to eliminate the mechanical coupling interference caused by the adjustment of clamping forces at adjacent stations include: Beforehand, mechanical coupling calibration of the workbench is performed. Under no-load conditions, a simulated clamping force with gradient changes is applied to each station one by one. The force change values ​​of the support points of all other stations are collected to construct a global coupling coefficient matrix, with matrix element K. ij This represents the coupling interference force experienced by the j-th workstation when a unit clamping force is applied to the i-th workstation; During real-time control, when a certain station needs to adjust the clamping force, the estimated value of the coupling interference of the clamping force adjustment on all other adjacent stations is calculated based on the global coupling coefficient matrix. The compensation adjustment is synchronously output to the support rods of all other adjacent workstations to counteract the influence of coupling interference force and ensure that the support force fluctuation of all non-adjustment workstations is less than the preset threshold.

[0075] Specifically, this algorithm addresses the mechanical coupling problem existing in multi-station shared worktables (i.e., the adjustment of clamping force in a single station is transmitted through the worktable structure, causing changes in the support force of other stations, which in turn leads to workpiece displacement, vibration, or even machining accuracy deviation). It achieves the adjustment of clamping force in multiple stations through a three-level logic of "pre-calibration modeling - real-time interference prediction - synchronous compensation cancellation".

[0076] (1) Pre-calibration stage: Construction of global coupling coefficient matrix (static modeling stage) The core of this step is to quantify the mechanical transmission characteristics of the workbench and establish a quantitative mapping relationship of interference between workstations, which is a preliminary foundational work for the algorithm.

[0077] Calibration logic: When the worktable is unloaded (without workpiece clamping), a gradient-changing simulated clamping force is applied to each station individually (covering the full range of the actual working of the fixture), while the force change values ​​of the corresponding support points are collected through the force sensors built into the support rods of all other stations.

[0078] Matrix Construction: Based on the collected load-response data, a global coupling coefficient matrix K with dimensions consistent with the number of workstations is constructed, where matrix elements K ijThis represents the magnitude of the coupling interference force experienced by the support point of the j-th workstation when a clamping force of 1 unit is applied to the i-th workstation. Each row of the matrix corresponds to the interference characteristics of a force-applying workstation on all other workstations, and each column corresponds to the superposition characteristics of the interference from all other workstations on a force-receiving workstation, thus fully quantifying the mechanical coupling transmission law of the worktable.

[0079] (2) Real-time calculation stage: generation of coupling interference prediction (dynamic sensing stage) This step operates in real time during the processing. Its core function is to quickly estimate the impact of any clamping force adjustment at a particular workstation on all other workstations. Triggering condition: When the system determines, based on the processing status perception results, that a certain station needs to increase / decrease the clamping force (e.g., if the cutting force increases, the clamping force needs to be increased to suppress tool deflection; if chatter is detected, the clamping force needs to be decreased to release stress), the system first locks the adjustment amount ΔF of the clamping force for this operation. i (i represents the workstation number to be adjusted).

[0080] Interference Calculation: Based on the pre-calibrated global coupling coefficient matrix, the estimated coupling interference ΔF of this adjustment amount on all other workstations j is quickly calculated through matrix operations. i→j =K ij ×ΔF i That is, the interference force experienced by each non-adjustment station is equal to the product of the corresponding coupling coefficient and the adjustment amount, thus realizing the advance quantification of the interference amount.

[0081] (3) Compensation execution stage: Synchronous adjustment to offset interference (closed-loop control stage) This stage is the execution end of the algorithm. It offsets the interference of the predicted value by synchronously outputting the compensation amount, thereby achieving force stability for non-adjustment workstations. Synchronous Compensation: While outputting the clamping force adjustment command to the adjustment station, a compensation adjustment amount ΔF, equal in magnitude but opposite in direction to the estimated value of the coupling interference, is simultaneously output to the support rods of all other non-adjustment stations. 补偿,j = -ΔF i→j It actively counteracts the coupling interference caused by the adjustment.

[0082] Performance Guarantee: The real-time response capability of the servo system ensures the synchronization of adjustment and compensation commands, ultimately ensuring that the actual fluctuation of the support force of all non-adjustment stations is less than the preset threshold (usually set to within 50N in precision machining scenarios), avoiding workpiece offset, vibration, or decreased machining accuracy caused by mechanical coupling between stations.

[0083] This algorithm is essentially a decoupled control strategy based on feedforward compensation. By understanding the mechanical transmission characteristics of the worktable in advance, it transforms the traditional feedback control of "correcting after interference occurs" into feedforward control of "actively canceling interference after prediction". This not only ensures the response speed of adjustment, but also avoids parameter coupling interference between multiple workstations, thus adapting to the efficiency and accuracy requirements of multi-workstation parallel processing.

[0084] In one possible embodiment, the digital model established in step S1 includes workpiece material properties, fixture support point and clamping point position parameters, and worktable mechanical coupling coefficient.

[0085] Specifically, the integrated digital model of workpiece-fixture-machine tool table constructed in this step is the core digital carrier for realizing virtual pre-clamping and automatic posture correction. The functions of the three types of parameters are as follows: Workpiece material property parameters: These parameters map the physical and mechanical properties of the workpiece, including core parameters such as elastic modulus, Poisson's ratio, and yield strength. They provide a material mechanics basis for deformation prediction and contact stress judgment during subsequent posture correction, and prevent local plastic deformation of the workpiece due to excessive support force during the lifting process.

[0086] The position parameters of the fixture support point and clamping point: map the geometric layout characteristics of the clamping system, including the spatial coordinates of the flexible support rods at each station, the arrangement position and stroke range of the clamping unit, clarify the motion boundary and action position of the clamping actuator, and establish the geometric mapping relationship between the "change in the lifting amount of the support rod" and the "change in the workpiece posture", which are the basic constraints for inverse kinematics solution.

[0087] Workbench mechanical coupling coefficient: Maps the mechanical transmission characteristics of a multi-station shared workbench. The coupling coefficient matrix obtained through pre-calibration quantifies the interference of single-station clamping force / support force adjustment on other stations. The mechanical coupling effect between stations is considered in advance during the initial lifting calculation stage to avoid the workpiece displacement of adjacent stations caused by single-station pose correction, and to ensure the overall consistency of multi-station clamping.

[0088] This model includes both geometric position matching and mechanical characteristic mapping, providing a digital environment close to real working conditions for accurate calculation of the initial lifting amount.

[0089] In one possible embodiment, the feature parameters extracted in step S3 include: the time-domain amplitude of the cutting force, the chatter characteristic frequency of the vibration signal in the frequency domain, the spindle power change rate, and the dynamic fluctuation of the force on the force-measuring support rod.

[0090] Specifically, the time-domain amplitude of the cutting force: Cutting force is a three-dimensional dynamic force generated during the cutting process when the tool contacts the workpiece. Its magnitude directly reflects the intensity of the current cutting load. The cutting force value in the time domain can be obtained by indirectly converting it through the force sensor built into the force support rod or the spindle current. By extracting its peak value, effective value and other amplitude characteristics, the load change during the cutting process can be directly determined.

[0091] When the tool enters the large-margin cutting area, the tool chipes, or the workpiece develops local hard spots, the time-domain amplitude of the cutting force will increase abnormally. The system can quickly identify the load change based on this feature and determine whether it is necessary to increase the clamping force to suppress workpiece deformation.

[0092] Flutter characteristic frequency in the frequency domain of vibration signal: Machining chatter is an unstable self-excited vibration of the tool-workpiece-fixture system. Its vibration signal has a specific characteristic frequency range (usually the coupling frequency of the cutting harmonic and the system's natural frequency). By performing a Fast Fourier Transform (FFT) on the time-domain vibration signal collected by the table accelerometer to convert it to the frequency domain, the amplitude characteristics of the corresponding chatter frequency range can be extracted, enabling early identification of chatter.

[0093] In the early stages of chatter, it may not be directly reflected in obvious changes in cutting force or power, but it can cause chatter marks on the machined surface and reduce tool life. This characteristic can provide early warning in the bud stage of chatter, and the system can promptly reduce the clamping force of the corresponding station to release internal stress and adjust the support damping to suppress vibration propagation.

[0094] Spindle power change rate: Spindle power is linearly and positively correlated with cutting load. The power change rate is the amplitude of spindle power change per unit time, reflecting the speed of change in cutting load. When the workpiece undergoes localized tool deflection, the actual cutting depth of the tool decreases, and the spindle power will experience an abnormal drop, with a rate of change much greater than the power fluctuations during normal cutting.

[0095] Compared to static power values, the rate of change characteristic can more sensitively identify the workpiece deflection trend and provide early warning before workpiece deformation leads to dimensional deviation. The system can simultaneously increase the support force and clamping force to suppress further deflection and ensure machining dimensional accuracy.

[0096] Dynamic fluctuation of force on the force-measuring support rod: The force-measuring support rod is in direct contact with the bottom surface of the workpiece, and the dynamic fluctuation of its force value reflects the actual stress and deformation state of the workpiece during the cutting process. During normal and stable cutting, the support force fluctuation is within a stable range; when the workpiece becomes loose, partially suspended, or undergoes uneven deformation, the support force will exhibit abnormal fluctuations.

[0097] This feature directly reflects the support stability of the clamping system and can be used to determine whether there is a gap in the initial clamping, whether the multi-station coupling adjustment interferes with the support of this station, and is a direct feedback indicator of the clamping force adjustment effect, ensuring that the clamping system is always in a rigid support state.

[0098] The four types of characteristic parameters cover the core abnormal modes of the processing from four dimensions: load intensity, vibration characteristics, load change rate, and support stability. Together, they constitute a quantitative perception system of the processing status, providing accurate judgment input for subsequent adaptive adjustment.

[0099] In one possible embodiment, the pose compensation after B-axis rotation in step S5 only applies to the overall offset of the workpiece, without adjusting the position of the support push rod. After the compensation is completed, the clamping force is replanned directly based on the cutting force parameters of the new machining surface.

[0100] Specifically, this mechanism is an efficient pose compensation strategy for B-axis face-changing machining scenarios in the multi-station flexible clamping adaptive control method of horizontal machining centers. The core logic is "one-time clamping and positioning, only coordinate compensation for face changing, no need for secondary mechanical adjustment". It replaces traditional mechanical position correction with software compensation of geometric offset to achieve seamless connection of multi-face machining.

[0101] During B-axis rotation, the workpiece is always rigidly contacted by the support rod with a slight force constraint, without any local relative displacement. Therefore, the pose deviation is only the overall six-degree-of-freedom offset of the workpiece (translation in the X / Y / Z directions and rotation around the axis), which is a holistic geometric offset. There is no situation requiring adjustment of the support rod position due to support point failure or local deformation of the workpiece. This offset can be fully quantified by acquiring the coordinates of the workpiece reference point through a pose sensor, without the need for mechanical adjustment of the support system.

[0102] In the initial clamping stage (step S1), the optimal configuration of the support rod positions has been achieved through inverse kinematics solution. All support rods achieve gapless surface contact with the bottom surface of the workpiece and meet the clamping accuracy threshold requirements. This support layout is adapted to the stiffness requirements of the entire machined surface of the workpiece. The B-axis rotation only changes the machining orientation of the workpiece and does not change the contact state between the workpiece and the support rods or the distribution of support stiffness. Therefore, the mechanical position of the support system does not need to be readjusted, and the original support constraints can be directly inherited to the machining process after the surface change.

[0103] The wall thickness and stiffness distribution of different machined surfaces vary significantly, and the cutting force parameters (magnitude, direction, and position of action) of the new process are completely different from those of the previous machined surface. The original clamping force parameters are no longer suitable for the new machining scenario. Therefore, after completing the coordinate system offset compensation, the initial clamping force planning algorithm (step S2) is directly called to recalculate the optimal clamping force threshold based on the stiffness characteristics and cutting force parameters of the new machined surface. This achieves accurate matching between the clamping parameters and the current machining process, avoiding problems such as deformation and tool deflection after surface change.

[0104] Compared to the traditional method of re-aligning and adjusting supports after changing surfaces, this mechanism can greatly reduce the auxiliary time for changing surfaces, while avoiding the accumulation of clamping errors caused by repeated mechanical adjustments, thus meeting the efficiency requirements of flexible production of multiple varieties and small batches.

[0105] Example 2, see Figure 6 The present invention also provides a multi-station flexible clamping adaptive control system for a horizontal machining center, comprising: a position sensor, a multi-axis flexible fixture execution module, a multi-source signal acquisition module, and an adaptive control terminal. The pose sensors are distributed at each station of the worktable to obtain the six-degree-of-freedom pose deviation of the workpiece at the initial stage and after the B-axis rotation. The multi-axis flexible clamping execution module includes arrayed servo-electric support rods and servo clamping units, which can independently control the position and force of each support point and clamping point; A multi-source signal acquisition module is integrated into the spindle, fixture, and worktable for real-time acquisition of current, force, and vibration signals. The adaptive control terminal, which incorporates a stiffness matching algorithm and a multi-station decoupling control algorithm, communicates with the posture sensor, the multi-axis flexible fixture execution module, the multi-source signal acquisition module, and the machine tool CNC system to execute the aforementioned adaptive control methods.

[0106] In one possible embodiment, the servo-electric support rod has a built-in force sensor and damping adjustment unit, which can continuously adjust the support force in the range of 0.1kN-10kN, and the damping parameters can be automatically adjusted according to the flutter characteristics.

[0107] In one possible embodiment, the pose sensor uses a non-contact laser displacement sensor array, which can complete the measurement of the workpiece pose within 1 second after the B-axis rotation stops, with a measurement accuracy of not less than 5μm.

[0108] In one possible embodiment, the adaptive control terminal communicates with the machine tool CNC system via a real-time bus, and the response delay for clamping force adjustment does not exceed 10ms, enabling dynamic and continuous adjustment during the machining process.

[0109] Specifically, this system is the physical hardware carrier of the above-mentioned adaptive control method. It adopts a closed-loop control architecture of "perception-decision-execution-feedback" and realizes automated and adaptive control of multi-station clamping process through the collaborative work of four core modules.

[0110] Pose sensor: It adopts a non-contact laser displacement sensor array, which is distributed at the edge of each station on the workbench, with a measurement accuracy of no less than 5μm and a fast response speed.

[0111] Initial clamping stage: Collect the actual three-dimensional coordinates of at least three non-collinear reference points on the workpiece surface, and perform point set matching with the theoretical coordinates of the corresponding reference points in the digital model (using the singular value decomposition (SVD) algorithm to solve the optimal matching matrix). Calculate the pose deviation of the workpiece relative to the theoretical clamping position in six degrees of freedom: X / Y / Z translation and A / B / C rotation, providing input for subsequent initial lifting calculation.

[0112] B-axis face-changing stage: The workpiece pose retest is completed within 1 second after the B-axis rotation stops, the overall offset of the workpiece after face-changing is quickly calculated, and the feedback is directly fed back to the CNC system to complete the automatic coordinate system compensation, replacing the traditional manual dial indicator alignment process.

[0113] Multi-axis flexible fixture execution module: Each workstation is equipped with an array of servo-electric support rods and servo clamping units. The support rods have built-in force sensors and damping adjustment units, which can continuously adjust the support force within the range of 0.1kN-10kN, and the damping parameters can be dynamically adjusted.

[0114] Support end: Receives lifting instructions from the adaptive control terminal and independently adjusts the extension height of each support rod to achieve precise correction of the workpiece's position and posture; during processing, it dynamically adjusts the support force and damping parameters according to control instructions to adapt to the stiffness requirements under different cutting conditions.

[0115] Clamping end: Based on the initial clamping force planning results and real-time adjustment commands, the output clamping force of each clamping unit is independently controlled to achieve dynamic matching between clamping force and workpiece rigidity and cutting load, avoiding excessive clamping force leading to workpiece deformation or insufficient clamping force leading to loosening during machining.

[0116] Multi-source signal acquisition module: The signal acquisition nodes are integrated in three types of locations: the spindle end acquires current / power signals, the clamp support rod has a built-in force sensor to acquire force signals, and the worktable surface is equipped with an accelerometer to acquire vibration signals.

[0117] During the machining process, three types of time-domain signals are simultaneously acquired and converted to the time-frequency domain to extract four core feature parameters: cutting force time-domain amplitude (reflecting load intensity), chatter characteristic frequency in the frequency domain of vibration signal (identifying chatter risk), spindle power change rate (judging tool deflection trend), and dynamic fluctuation of support force (monitoring clamping stability). This enables the quantitative characterization of three typical machining anomalies: sudden changes in cutting force, chatter, and local tool deflection, providing data input for adaptive control decisions.

[0118] Adaptive control terminal: It adopts an industrial-grade real-time controller, with embedded stiffness matching algorithm, multi-station decoupling control algorithm and digital model. It communicates with all sensors, execution modules and machine tool CNC system through real-time bus, and the command response delay does not exceed 10ms.

[0119] Pre-planning stage: Integrate workpiece CAD model, process parameters, worktable coupling coefficient and other data to establish a digital model of "workpiece-fixture-worktable", output the initial lifting amount through inverse kinematics solution, output the initial clamping force threshold through finite element analysis and multi-objective optimization solution, and complete the pre-configuration of clamping parameters.

[0120] Real-time control phase: Based on the characteristic parameters input by the multi-source signal acquisition module, the machining state is determined and a differentiated adjustment strategy is executed: when the cutting force increases, the clamping force and support force are increased simultaneously to suppress deformation; when chatter is detected, the clamping force is reduced and the support damping is adjusted to suppress vibration; at the same time, based on the pre-calibrated global coupling coefficient matrix, when the clamping force is adjusted at a single station, the compensation amount is output to the other stations simultaneously to eliminate mechanical coupling interference between stations.

[0121] Face-changing adaptation stage: Receive the pose deviation data after B-axis rotation, send pose compensation commands to the CNC system, and re-plan the clamping force based on the stiffness and cutting force parameters of the new machining surface to achieve seamless connection of multi-face machining.

[0122] This system constructs a closed-loop control logic of "virtual pre-clamping - initial stiffness matching - machining status perception - dynamic decoupling adjustment - automatic compensation for face change": During initial clamping, the position sensor collects data and uploads it to the control terminal. The terminal outputs the lifting amount of the push rod and the initial clamping force command, and the execution module completes automatic alignment and pre-clamping. During the processing, the multi-source signal acquisition module provides real-time feedback on the processing status, and the terminal dynamically adjusts the clamping parameters based on the built-in algorithm, while simultaneously completing multi-station decoupling compensation. When changing surfaces along the B-axis, the position sensor quickly remeasures the offset, and the terminal completes CNC coordinate compensation and clamping force replanning. It can then enter the adaptive control process for the next machining surface without manual intervention.

[0123] Compared to traditional rigid clamping systems, this system can significantly reduce machining deformation of thin-walled parts, significantly improve multi-station clamping efficiency, greatly shorten B-axis face-changing and alignment time, and effectively suppress machining chatter and tool deflection problems, making it suitable for flexible production scenarios with multiple varieties and small batches.

[0124] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A horizontal machining center multi-station flexible clamping adaptive control method, characterized in that, Includes the following steps: The CAD model and theoretical clamping position of the workpiece at each station are obtained to establish a digital model. The actual position and posture of the workpiece are obtained, the position and posture deviation is calculated, and the initial lifting amount of the flexible support top rod at each station is generated. Based on the CAD model, extract the wall thickness and stiffness distribution information of the workpiece at each station. Combined with the pre-estimated cutting force, use finite element analysis to calculate the deformation of the workpiece under different support and clamping points. Then, with the minimum deformation as the objective function, solve the initial clamping force threshold of the clamping unit at each station. During the machining process, the spindle current and power signals, the force signal of the force support rod, and the acceleration signal of the worktable are collected in real time, and time-frequency analysis is performed to extract characteristic parameters. Based on the extracted feature parameters, the current processing status of each station is determined and the clamping force is adjusted. The multi-station decoupling control algorithm is executed to eliminate the mechanical coupling interference caused by the clamping force adjustment of adjacent stations. When the B-axis rotates for surface changing, the clamping force is paused while maintaining micro-force support. The workpiece offset is then reacquired and fed back to the CNC system for pose compensation. The clamping force is then replanned on the new machining surface.

2. The method of claim 1, wherein, When adjusting the clamping force, if the cutting force at a certain station increases and there is a risk of tool deflection, the clamping force and support force at that station are increased proportionally; if chatter is detected at a certain station, the clamping force at that station is reduced to release stress, and the stiffness damping of the support point is adjusted.

3. The method of claim 1, wherein, When generating the initial lift, the following are included: The actual three-dimensional coordinates of at least three non-collinear reference points on the workpiece surface are collected by a pose sensor and matched with the theoretical coordinates of the corresponding reference points in the digital model to calculate the six-degree-of-freedom pose deviation of the workpiece relative to the theoretical clamping position. Based on the six-degree-of-freedom pose deviation, and taking the lifting range of each support rod and the contact constraint between the rod and the bottom surface of the workpiece as boundary conditions, an inverse kinematics solution model for workpiece pose correction is established. The initial lifting amount of each flexible support rod is obtained by solving, so that the positional deviation of the workpiece after correction is less than the preset clamping accuracy threshold, and all support rods maintain surface contact with the bottom surface of the workpiece without gaps. In the inverse kinematics solution model, a worktable flatness error compensation parameter is introduced; the clamping accuracy threshold is dynamically set according to the machining process accuracy requirements of the corresponding workstation.

4. The method of claim 1, wherein, When calculating deformation, the following should be included: Import the CAD model of the workpiece and refine the mesh in the thin-walled area, clamping contact area, and cutting area. The mesh size is 1 / 3 to 1 / 5 of that in non-critical areas. In the finite element model, a fixed displacement constraint is applied to the position of the support rod, a normal pressure load is applied to the position of the clamping unit, and a three-dimensional dynamic cutting force load is applied to the cutting position of the machining process. Static simulation is performed to extract the maximum normal deformation of the workpiece's machined surface, the stress distribution in the clamping contact area, and the reaction force data of the support points as the deformation calculation results.

5. The method of claim 1, wherein, Solving for the initial clamping force threshold includes: A multi-objective optimization function is constructed, with the core objective being to minimize the maximum normal deformation of the workpiece's machined surface, and the auxiliary objectives being to minimize the total clamping force and the stress difference between each clamping point. Set constraints: including the clamping force adjustment range constraint of each clamping unit, the stress constraint that the maximum stress in the workpiece contact area does not exceed 70% of the material yield strength, and the full contact constraint that the reaction force of all support points is greater than 0. The objective function is solved iteratively using a particle swarm optimization algorithm, and the optimal initial clamping force threshold for each clamping unit is output.

6. The method of claim 1, wherein, When executing the multi-station decoupling control algorithm, the following are included: In the idle state, the analog clamping force of each station is changed one by one, the force change values of all the remaining station support points are collected, the global coupling coefficient matrix is constructed, and the matrix element K ij represents the coupling interference force suffered by the jth station when the ith station applies a unit clamping force. During real-time control, when a workstation needs to adjust the clamping force, the estimated value of the coupling interference of the clamping force adjustment on all other adjacent workstations is calculated based on the global coupling coefficient matrix. The compensation adjustment is synchronously output to the support rods of all other adjacent workstations to counteract the influence of coupling interference force.

7. The method of claim 3, wherein, The digital model includes workpiece material properties, fixture support point and clamping point position parameters, and worktable mechanical coupling coefficient.

8. The method of claim 1, wherein, The characteristic parameters include the time-domain amplitude of the cutting force, the chatter characteristic frequency of the vibration signal in the frequency domain, the spindle power change rate, and the dynamic fluctuation of the force on the force-measuring support rod.

9. The method according to claim 1, characterized in that, Posture compensation after B-axis rotation only applies to the overall workpiece offset, without needing to adjust the position of the support push rod. After compensation, the clamping force is recalculated based on the cutting force parameters of the new machining surface.

10. A multi-station flexible clamping adaptive control system for a horizontal machining center, characterized in that, include: The pose sensors are distributed at each station of the worktable to obtain the six-degree-of-freedom pose deviation of the workpiece at the initial stage and after the B-axis rotation. The multi-axis flexible clamping execution module includes arrayed servo-electric support rods and servo clamping units, which can independently control the position and force of each support point and clamping point; A multi-source signal acquisition module is integrated into the spindle, fixture, and worktable for real-time acquisition of current, force, and vibration signals. An adaptive control terminal, with embedded stiffness matching algorithm and multi-station decoupling control algorithm, is communicatively connected to a posture sensor, a multi-axis flexible fixture execution module, a multi-source signal acquisition module, and a machine tool CNC system, respectively, and is used to execute the adaptive control method as described in any one of claims 1-9.

Images