Process parameter control method and system applied to fastener production

Abstract

The invention relates to the technical field of fastener process production, in particular to a process parameter control method and system applied to fastener production. The method comprises the following steps: acquiring three-dimensional cutting force of a tool-workpiece contact area in a fastener thread machining process in real time, and generating cutting force dynamic fluctuation characteristic data; the real-time abnormal situation of the current machining process is recognized according to the cutting force dynamic fluctuation characteristic data, a tool wear mode is judged, thread ring machining deviation prediction is conducted, and a predicted thread geometric deviation value is generated; and adjusting multiple machining control parameter values according to the predicted thread geometric deviation value, and carrying out real-time machining feedback monitoring so as to realize fastener production self-adaptive control. According to the method, intelligent recognition of the tool wear mode and accurate prediction of the machining deviation are achieved through cutting force sensing, and it is ensured that the thread machining precision of the fastener is stable and controllable through multi-parameter cooperative self-adaptive regulation and control.

Description

Technical Field

[0001] The present invention relates to the technical field of fastener production technology, and in particular to a method and system for controlling process parameters of fastener production. Background Art

[0002] Tool wear is a common and unavoidable phenomenon during fastener thread machining. As machining time progresses, the tool surface gradually wears, causing changes in its geometry. This progressive wear is neither linear nor predictable; its rate and pattern are complexly influenced by factors such as the workpiece material, cutting parameters, and cooling conditions. It is this progressive tool wear that has become one of the major bottlenecks affecting thread machining accuracy and surface quality. Especially when machining high-precision fine threads, due to the more stringent requirements for geometric dimensions and form and position tolerances, even slight tool wear can lead to serious problems such as cumulative pitch error, changes in effective thread diameter, and deviations in thread profile angle. These deviations not only directly reduce the thread's geometric accuracy but also significantly deteriorate the thread's surface quality, such as burrs, scratches, and increased surface roughness, which in turn impact the fatigue life and reliability of the fastener. Currently, traditional methods for controlling process parameters in fastener production rely primarily on empirical rules, preset machining parameters, periodic offline testing, or post-processing corrections based on statistical process control (SPC). These methods can guarantee product qualification rates to a certain extent, but their limitations are becoming increasingly prominent. For example, offline inspection is usually performed after machining is completed, which has a lag effect. This means that it can only be discovered when thread deviations have already occurred. At this point, a batch of unqualified products has already been produced, resulting in high scrap rates and rework costs. Summary of the Invention

[0003] Based on this, the present invention provides a method and system for controlling process parameters in fastener production to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a method for controlling process parameters in fastener production is provided, comprising the following steps:

[0005] Step S1: Real-time acquisition of the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining, dynamic trajectory tracking, and generation of three-dimensional vector trajectory data of the cutting force; thread profile feature analysis based on the three-dimensional vector trajectory data of the cutting force to generate thread profile cutting force feature data; correlation of thread turn fluctuation amplitude based on the thread profile cutting force feature data to generate dynamic fluctuation feature data of the cutting force;

[0006] Step S2: performing a cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data;

[0007] Step S3: Scan the contour of the local area of ​​the machined thread surface of the fastener, calculate the local deviation value of the thread, and generate the local precision deviation value of the thread; judge the tool wear pattern based on the real-time abnormal machining status data and the local precision deviation value of the thread, and predict the thread circle machining deviation to generate the predicted thread geometry deviation;

[0008] Step S4: Adjust multiple processing control parameter values ​​according to the predicted thread geometry deviation to generate a composite thread processing adjustment instruction; issue a real-time execution control instruction to the machine tool based on the composite thread processing adjustment instruction, and perform real-time processing feedback monitoring to achieve adaptive control of fastener production.

[0009] Preferably, the present invention further provides a process parameter control system for fastener production, which executes the process parameter control method for fastener production as described above. The process parameter control system for fastener production includes:

[0010] The fastener cutting force analysis module is used to collect the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining in real time, and dynamically track the trajectory to generate three-dimensional cutting force vector trajectory data; based on the three-dimensional cutting force vector trajectory data, it analyzes the thread profile characteristics to generate thread profile cutting force characteristic data; based on the thread profile cutting force characteristic data, it correlates the thread circle fluctuation amplitude to generate cutting force dynamic fluctuation characteristic data;

[0011] The machining anomaly monitoring module is used to perform vibration suppression analysis during the cutting process based on the dynamic fluctuation characteristic data of the cutting force, so as to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data;

[0012] The accuracy deviation prediction module is used to scan the local area profile of the machined thread surface of the fastener, calculate the local thread deviation value, and generate the local thread accuracy deviation value; based on the real-time abnormal processing status data and the local thread accuracy deviation value, it determines the tool wear pattern and predicts the thread circle processing deviation to generate the predicted thread geometry deviation;

[0013] The adaptive production control module is used to adjust the values ​​of multiple processing control parameters based on the predicted thread geometry deviation and generate compound thread processing adjustment instructions; based on the compound thread processing adjustment instructions, the machine tool executes control instructions in real time and performs real-time processing feedback monitoring to achieve adaptive control of fastener production.

[0014] The present invention generates three-dimensional cutting force vector trajectory data by real-time acquisition of three-dimensional cutting forces in the tool-workpiece contact zone during fastener thread machining and performing precise dynamic tracking of these forces. This multi-dimensional, high-resolution, real-time data acquisition method reveals the dynamic and fluctuating mechanical characteristics of the thread profile during its formation process with unprecedented depth, thereby generating precise thread profile cutting force characteristic data and cutting force dynamic fluctuation characteristic data. This enables the system to accurately and early identify minor vibrations and abnormal conditions caused by factors such as tool wear, unstable cutting, or material unevenness during the cutting process, effectively avoiding the inherent lag inherent in traditional offline detection and fundamentally preventing the mass production of substandard products. Vibration suppression analysis during the cutting process based on the dynamic fluctuation characteristic data can identify real-time abnormal conditions during the current machining process. This means that the system can detect subtle deteriorations in the machining state, such as abnormal fluctuations in cutting force and increased vibration, before thread geometry deviations become noticeable. These are early warning signals of tool wear or changes in cutting conditions, rather than waiting until problems accumulate to an irreversible level. By scanning the contours of local areas of the machined thread surface and calculating local thread accuracy deviations, combined with real-time abnormal machining data, the system can conduct in-depth analysis of tool wear patterns. This analysis, combining process data with local result data, not only determines tool wear but also identifies specific wear patterns (such as flank wear, rake wear, or chipping), a feat difficult to achieve with traditional methods. More importantly, based on wear pattern analysis, the system can predict machining deviations for subsequent thread turns and generate predicted thread geometry deviations. This predictive capability is key to enabling proactive control, transforming a passive response to tool wear into a proactive assessment of thread geometry deviations, significantly reducing scrap and rework costs. Based on the predicted thread geometry deviations, the system intelligently adjusts multiple machining control parameters, including but not limited to feed rate, depth of cut, spindle speed, and even cooling and lubrication conditions, generating complex thread machining adjustment commands. These commands are issued to the machine tool for real-time execution, and real-time machining feedback monitoring ensures the effectiveness of the adjustments. This closed-loop, adaptive control strategy enables the production process to dynamically respond to the gradual changes in tool wear and random fluctuations in cutting conditions, fundamentally suppressing high-precision fine-thread processing difficulties such as cumulative pitch error, changes in effective thread diameter, and tooth profile angle deviation. Therefore, the present invention's method for controlling process parameters in fastener production, which includes full-chain intelligent control from cutting force perception, processing situation identification, geometric deviation prediction to parameter adaptive adjustment, completely solves the pain point of traditional methods in achieving sustained and stable thread accuracy under the background of tool wear. It not only greatly improves the processing accuracy and surface quality of fasteners, but also significantly extends the service life of tools and reduces production costs and dependence on skilled operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic flow chart of the steps of the method for controlling process parameters in fastener production according to the present invention;

[0016] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0017] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0018] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0019] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0020] To achieve this, please refer to Figure 1 The present invention provides a method for controlling process parameters in fastener production, comprising the following steps:

[0021] Step S1: Real-time acquisition of the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining, dynamic trajectory tracking, and generation of three-dimensional vector trajectory data of the cutting force; thread profile feature analysis based on the three-dimensional vector trajectory data of the cutting force to generate thread profile cutting force feature data; correlation of thread turn fluctuation amplitude based on the thread profile cutting force feature data to generate dynamic fluctuation feature data of the cutting force;

[0022] Step S2: performing a cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data;

[0023] Step S3: Scan the contour of the local area of ​​the machined thread surface of the fastener, calculate the local deviation value of the thread, and generate the local precision deviation value of the thread; judge the tool wear pattern based on the real-time abnormal machining status data and the local precision deviation value of the thread, and predict the thread circle machining deviation to generate the predicted thread geometry deviation;

[0024] Step S4: Adjust multiple processing control parameter values ​​according to the predicted thread geometry deviation to generate a composite thread processing adjustment instruction; issue a real-time execution control instruction to the machine tool based on the composite thread processing adjustment instruction, and perform real-time processing feedback monitoring to achieve adaptive control of fastener production.

[0025] In an embodiment of the present invention, the method for controlling process parameters in fastener production includes the following steps:

[0026] Step S1: Real-time acquisition of the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining, dynamic trajectory tracking, and generation of three-dimensional vector trajectory data of the cutting force; thread profile feature analysis based on the three-dimensional vector trajectory data of the cutting force to generate thread profile cutting force feature data; correlation of thread turn fluctuation amplitude based on the thread profile cutting force feature data to generate dynamic fluctuation feature data of the cutting force;

[0027] In an embodiment of the present invention, a KISTLER 9257B high-frequency three-axis force sensor is installed on the fastener thread processing equipment, with the sampling frequency set to 10,000 Hz. The analog signal is converted into a digital signal using a KISTLER 5070A charge amplifier. Instantaneous data of radial, axial, and tangential cutting forces are collected and stored in a three-column time series matrix. The amplitude energy of the collected three-dimensional cutting force data is calculated, and the square root of the sum of the squares of the components in the formula is taken to generate synthetic cutting force data. A coordinate system with the thread axis as the Z axis is established using a three-dimensional spatial mapping algorithm. The cutting force data is matched with the tool position to generate cutting force spatial distribution data. The cutting force density in the tooth top fillet area and the cutting force components on the tooth flank are extracted for geometric coupling analysis. The cutting force characteristic data is segmented into periodic data according to the standard pitch spacing, and the average and peak values ​​of each period are calculated. The dynamic fluctuation coefficient of the force ratio is obtained through processing. The peak envelope and mean trajectory are generated through spline interpolation to evaluate the stability of the cutting process and correlate the fluctuation amplitude of the thread turns to generate cutting force dynamic fluctuation characteristic data. The entire data collection and processing process is carried out in real time, and the processing delay is controlled within 1 millisecond.

[0028] Step S2: performing a cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data;

[0029] In an embodiment of the present invention, a multi-scale entropy analysis is performed based on the dynamic fluctuation characteristic data of the cutting force, using a sample entropy algorithm, with the embedding dimension set to 2 and the similarity tolerance set to 0.15 times the standard deviation. The entropy value at each scale is calculated to construct an entropy spectrum curve, and the spectrum characteristics extracted by Fourier transform are combined to form a comprehensive evaluation index of the vibration characteristics. A mapping relationship between the vibration characteristics and the tool wear amount is established by a support vector regression algorithm to generate a tool wear indicator factor. The benchmark average cutting force vector data and the benchmark cutting force peak-to-average ratio data of the new tool processing are obtained as reference standards for abnormality identification. The periodic average cutting force vector data obtained in real time are subtracted from the benchmark data to calculate the average force deviation vector data, and the deviation sequence of the force ratio dynamic fluctuation coefficient and the benchmark peak-to-average ratio is calculated. The normal area and the abnormal area are divided in the two-dimensional feature space, and the boundary of the normal area is that the average force deviation vector modulus is less than 0.15 and the peak-to-average ratio deviation root mean square value is less than 0.12. Density clustering analysis is performed on abnormal points beyond the boundary, and the DBSCAN algorithm is used to classify abnormal situations into three categories: abnormal tool wear, abnormal cutting parameters, and abnormal material properties. Real-time abnormal processing situation data is generated, and the data update frequency is 1 Hz.

[0030] Step S3: Scan the contour of the local area of ​​the machined thread surface of the fastener, calculate the local deviation value of the thread, and generate the local precision deviation value of the thread; judge the tool wear pattern based on the real-time abnormal machining status data and the local precision deviation value of the thread, and predict the thread circle machining deviation to generate the predicted thread geometry deviation;

[0031] In an embodiment of the present invention, a Keyence VR-6000 ultra-depth three-dimensional optical profiler is used to scan the machined thread surface of the fastener, with a vertical resolution of 0.1 microns and a horizontal resolution of 0.5 microns. The scanning range is set to 10 mm in the axial direction and covers an angle range of 120 degrees in the circumferential direction. The thread axis is determined by the RANSAC cylindrical fitting algorithm, and a cylindrical coordinate system with the axis as the Z axis is established. The tooth vertex set, tooth bottom point set and tooth profile middle point are extracted to calculate the instantaneous expansion and contraction of the pitch, the instantaneous radial offset of the effective diameter, the instantaneous distortion of the tooth profile angle and the instantaneous curvature change of the tooth top fillet radius. The measured data is compared with the national standard, the relative deviation is calculated, and the cumulative state of the pitch contraction, the state of the tooth profile angle distortion, the state of the effective diameter offset and the state of the tooth top fillet curvature deviation are judged. Dynamic weights are assigned to the four states, and the weight coefficients are calculated based on the degree of influence on the fastener assembly performance, the severity of the deviation and the coverage range to generate the local accuracy deviation value of the thread. Real-time abnormal machining data is combined to determine tool wear patterns. The degree of tool tip and side wear is estimated by measuring the curvature change of the thread crest radius and the effective diameter offset. The overall tool wear percentage is calculated, and a neural network prediction model is used to generate the predicted geometric deviation for the next thread turn, including thread diameter deviation, pitch deviation, and profile angle offset.

[0032] Step S4: Adjust multiple processing control parameter values ​​according to the predicted thread geometry deviation to generate a composite thread processing adjustment instruction; issue a real-time execution control instruction to the machine tool based on the composite thread processing adjustment instruction, and perform real-time processing feedback monitoring to achieve adaptive control of fastener production.

[0033] In this embodiment of the present invention, a compensation value for the spindle speed and feed axis synchronization ratio is calculated based on the predicted pitch deviation in the thread geometry deviation. The predicted pitch deviation is subjected to cumulative pitch error analysis to calculate the single-turn pitch compensation. The spindle speed and thread lead data are read from the CNC system, and the angular velocity adjustment is calculated. This is then decomposed into an instantaneous angular velocity adjustment sequence with each 30-degree adjustment point, generating a spindle instantaneous phase compensation instruction. The radial cutting depth adjustment value is calculated based on the thread circle diameter deviation, and the workpiece coordinate system offset is modified using the G10 L20 instruction to generate a cutting depth adjustment instruction. The coolant flow and pressure adjustment coefficients are calculated based on the thread profile angle offset, with a standard flow rate of 18 liters / minute and a standard pressure of 2.0 MPa, controlled by the M138 and M139 instructions, respectively. The feed rate fine-tuning amount is calculated based on the thread profile angle deviation and pitch deviation, and updated in real time using the G01 F instruction. All instructions are combined into a composite thread machining adjustment instruction, with execution priority and time sequence set. The instructions are then transmitted to the execution unit via the CNC system's built-in Ethernet interface. The system monitors the execution effect in real time and triggers the protection mechanism when the parameters exceed the safety range, realizing closed-loop adaptive control of the fastener thread processing process.

[0034] Preferably, in step S1, real-time acquisition of the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining and dynamic trajectory tracking includes:

[0035] The high-frequency three-dimensional force sensor installed on the machine tool turret is used to synchronously collect the radial cutting force, axial cutting force and tangential cutting force in the tool-workpiece contact area during the fastener thread machining process, and obtain instantaneous three-dimensional cutting force time series data;

[0036] The radial component, axial component and tangential component in the instantaneous three-dimensional cutting force time series data are calculated at each digital sampling point to obtain the instantaneous series data of the cutting force amplitude energy.

[0037] Perform square root operation on instantaneous series data of cutting force amplitude energy to obtain instantaneous synthetic cutting force data;

[0038] The three-dimensional vector trajectory of the cutting force is dynamically tracked according to the instantaneous synthetic cutting force data to generate the three-dimensional vector trajectory data of the cutting force.

[0039] In an embodiment of the present invention, a high-frequency three-dimensional piezoelectric quartz crystal cutting force sensor (e.g., Kistler 9129AA or equivalent) is securely mounted on the tool holder of a fastener threading machine tool to ensure that the sensor's measuring axis is precisely aligned with the radial, axial, and tangential directions of the machine tool's machining coordinate system. The sensor is connected to a high-frequency data acquisition card (e.g., NI PXIe-6358 or equivalent), which integrates a signal conditioning module to amplify, filter, and perform analog-to-digital conversion on the weak charge signal output by the sensor. The data acquisition system sets the sampling frequency to 20,000 Hz to ensure that transient force fluctuations during the cutting process are captured. When the fastener thread cutting process is started, the data acquisition system is started synchronously to continuously collect the radial cutting force, axial cutting force, and tangential cutting force components of the tool-workpiece contact area. The collected data is stored in the form of an instantaneous three-dimensional cutting force time series data stream corresponding to the timestamp and force value, and the data format is a discrete digital sequence. The instantaneous three-dimensional cutting force time series data collected in the previous step includes the timestamp corresponding to each digital sampling point and the radial, axial, and tangential cutting force values ​​at that moment. A data processing system, an industrial control computer equipped with a high-performance processor, receives this three-dimensional time series data stream. A dedicated computational module within this computer performs a specific mathematical operation on each received digital sampling point: squaring the radial, axial, and tangential cutting force values ​​recorded at that sampling point, and then summing these three squared values. This summation operation is performed independently and repeatedly at each discrete digital sampling point, generating a new time series. Each data point in this series represents the squared amplitude of the resultant cutting force in three-dimensional space at that instant. This series is the instantaneous cutting force amplitude energy series data. Each data point in the obtained instantaneous cutting force amplitude energy series data represents the squared amplitude of the three-dimensional resultant cutting force at the corresponding instant. In this step, the dedicated computational module of the industrial control computer continues to process these squared values. For each digital data point in the instantaneous series data of cutting force amplitude energy, the module performs an accurate square root operation. This square root operation restores each square value to its original, non-negative root value. Each data point of the operation result represents the scalar magnitude of the actual force acting in the tool-workpiece contact area at that instant, that is, the instantaneous composite cutting force. This process continuously iterates the entire instantaneous series data of cutting force amplitude energy, and finally generates a new instantaneous composite cutting force data sequence, which accurately reflects the change of cutting force amplitude over time during fastener thread processing. The original collected instantaneous three-dimensional cutting force time series data (that is, the values ​​of radial cutting force, axial cutting force and tangential cutting force at each sampling point) is used as the data input of the three-dimensional coordinate axis. A high-performance industrial control computer is connected to a visualization display.The computer's graphics processing unit maps the radial cutting force values ​​in the original three-dimensional cutting force time-series data to the X-axis of the three-dimensional coordinate system, the axial cutting force values ​​to the Y-axis, and the tangential cutting force values ​​to the Z-axis. At each digital sampling moment, the three components of the force constitute a coordinate point in three-dimensional space. As time passes, these consecutive coordinate points are connected, plotting a dynamic curve that evolves over time in the three-dimensional force space. The instantaneous composite cutting force data assigns the length information of the vector from the origin to that point to each point on this curve. The entire process is rendered in real time on a visual display, showing the dynamic evolution path of the cutting force vector in three-dimensional space. This is the three-dimensional cutting force vector trajectory data. This trajectory data contains the (radial force, axial force, tangential force) coordinate points at each moment, as well as the instantaneous composite cutting force value corresponding to that moment.

[0040] Preferably, the thread profile evaluation according to the cutting force three-dimensional vector trajectory data in step S1 includes:

[0041] Based on the three-dimensional vector trajectory data of cutting force, the fastener thread spatial area is mapped to generate the thread cutting force spatial distribution data;

[0042] According to the spatial distribution data of thread cutting force, the cutting force density of the thread top fillet radius area in the fastener is analyzed to obtain the top cutting force density data;

[0043] Extract the cutting force component of the thread tooth flank inclination angle area in the fastener based on the thread cutting force spatial distribution data to obtain the tooth flank cutting force component data;

[0044] The thread profile geometry coupling analysis is performed based on the tooth top cutting force density data and the tooth side cutting force component data to obtain the thread forming mechanical distribution data;

[0045] Obtain fastener effective diameter size constraint data;

[0046] The radial force concentration analysis of the thread forming mechanical distribution data is carried out through the effective diameter size constraint data of the fastener, and then the thread tooth profile angle forming mechanical evaluation is carried out to generate the thread tooth profile cutting force characteristic data.

[0047] In an embodiment of the present invention, a cylindrical coordinate system is established with the thread axis as the Z axis and the entry point as the origin, and the coordinate system resolution is 0.01 mm. The three-dimensional cutting force vector F(k) (Fx(k), Fy(k), Fz(k)) is synchronously matched with the corresponding tool position data P(k) (Px(k), Py(k), Pz(k)), and the matching accuracy is 0.001 milliseconds. Through the spatial interpolation algorithm, the radial basis function interpolation method is adopted, the interpolation kernel function uses the Gaussian function, the influence radius is set to 0.5 mm, and a 100×60×360 three-dimensional grid point is constructed on the fastener thread geometric model, and the grid point spacing is 0.05 mm. The cutting force vector value is calculated for each grid point to generate the thread cutting force spatial distribution data. The data is stored as a three-dimensional tensor with a tensor size of 100×60×360×3, where the first three dimensions represent the spatial position and the last dimension represents the three-dimensional cutting force vector at that position. Based on the standard geometric parameters of the thread tooth profile, the tooth top fillet radius area is determined. For M10×1.5 fasteners, the tooth top fillet radius is 0.144 mm. A data subset of the tooth top fillet region is extracted from the thread cutting force spatial distribution data. This extraction method sets a spherical query radius of 0.15 mm and performs a spherical region query centered on the tooth top circle in the standard tooth profile geometry model. Within the extracted region, the cutting force per unit area is calculated using the formula: ρ(i) = |F(i)| / A(i), where ρ(i) represents the cutting force density at the i-th grid point, in Newtons per square millimeter; |F(i)| represents the modulus of the cutting force vector at that point, in Newtons; and A(i) represents the area represented by that grid point, in square millimeters, with a value of 0.0025 square millimeters. The cutting force density distribution in the tooth top region is statistically analyzed to generate a heat map of the tooth top cutting force density with a resolution of 0.02 mm. The thread tooth flank inclination angle region is determined. For M10×1.5 fasteners, the flank inclination angle is 30 degrees. In the spatial distribution data of thread cutting forces, the tooth flank area is located using a normal screening algorithm. The screening criteria are regional points where the angle between the surface normal vector and the horizontal plane is between 27 and 33 degrees. For the selected tooth flank area points, a local coordinate system is established, with the x-axis along the tangential direction of the tooth flank, the y-axis along the normal direction of the tooth flank, and the z-axis along the axial direction of the thread. The cutting force vector F(i) (Fx(i), Fy(i), Fz(i)) in the global coordinate system is transformed to the local coordinate system using a coordinate transformation matrix to obtain the local cutting force components F'(i) (Fx'(i), Fy'(i), Fz'(i)). Among them, Fy'(i) represents the normal cutting force component perpendicular to the tooth flank, in Newtons; Fx'(i) and Fz'(i) represent the cutting force components along the tangential and axial directions of the tooth flank, respectively, in Newtons. A tooth flank cutting force component distribution map is generated with a spatial resolution of 0.01 mm. A tooth geometry model is established with a resolution of 0.005 mm, including key areas such as tooth top, tooth side, and tooth bottom.The coupling relationship matrix C between the cutting force and the tooth profile's geometric features was calculated. The matrix has 6×6 dimensions and represents the degree of coupling between the six geometric feature points and the six mechanical parameters. The coupling coefficient is calculated as follows: C(m,n) = w(m)·v(n)·r(m,n), where C(m,n) represents the dimensionless coupling coefficient between the mth geometric feature point and the nth mechanical parameter; w(m) represents the weight coefficient of the mth geometric feature point, ranging from 0.6 to 1.0; v(n) represents the sensitivity coefficient of the nth mechanical parameter, ranging from 0.7 to 1.0; and r(m,n) represents the correlation coefficient between the two, ranging from -1.0 to 1.0, obtained through Pearson correlation analysis. Based on the coupling matrix, thread forming mechanical distribution data are generated. The data format is a two-dimensional matrix, with the number of rows corresponding to the number of sampling points along the tooth profile and the number of columns corresponding to the number of mechanical parameters. Measurements were made using the three-line method, with a force of 0.5 Newton and a measurement location at the center of the thread, with four measurement points uniformly sampled along the circumference. The measured effective diameter values ​​were compared with the standard values, and the effective diameter deviation, Δd, was calculated in millimeters. For M10×1.5 fasteners, the standard effective diameter is 9.026 mm, the tolerance grade is 6g, the upper deviation is -0.022 mm, and the lower deviation is -0.104 mm. The measurement system repeatability is ±0.003 mm. Based on the measurement results, effective diameter constraints were established. The constraint data consisted of the nominal diameter value, actual diameter value, deviation value, and diameter distribution data at four circumferential measurement points, forming an effective diameter constraint dataset. The dataset format was a structured array containing eight numeric fields. The fastener effective diameter constraint data was combined with thread forming mechanical distribution data to analyze radial force concentration. The radial force concentration index K is calculated using the following formula: K = ∑(Fr(i)·w(i)) / ∑Fr(i), where K represents the radial force concentration index, which is dimensionless; Fr(i) represents the radial cutting force value at the i-th radial sampling point, in Newtons; and w(i) represents the weight coefficient of the point, which is inversely proportional to the distance from the point to the effective diameter, in 1 / mm. The correlation between the thread profile angle deviation and the cutting force distribution is calculated using the grey correlation analysis method, with the correlation threshold set at 0.75. When the correlation is higher than the threshold, it is considered that there is a significant correlation between the abnormal cutting force distribution and the thread profile angle deviation. Based on the radial force concentration index and the thread profile angle correlation, the thread profile quality score Q is calculated, with a score range of 0 to 100. The scoring formula takes into account the effective diameter deviation, radial force concentration, and angle deviation factors to generate a thread profile cutting force characteristic data report, which includes the thread profile geometry parameters, cutting force distribution characteristics, and quality score.

[0048] Preferably, in step S1, correlating the thread circle fluctuation amplitude according to the thread profile cutting force characteristic data includes:

[0049] The thread profile cutting force characteristic data is periodically segmented according to the preset fastener thread pitch standard spacing to obtain the single pitch cutting force cycle data;

[0050] According to the single pitch cutting force cycle data, the average value of each direction force in each thread pitch processing cycle is calculated to obtain the cycle average cutting force vector data;

[0051] According to the single pitch cutting force cycle data, the peak values ​​of the forces in each direction in each thread pitch processing cycle are extracted to obtain the cycle peak cutting force vector data;

[0052] The cycle peak cutting force vector data is processed into a dynamic peak envelope, and the cycle average cutting force vector data is processed into a dynamic mean trajectory. Then, the ratio of the instantaneous cutting force peak to the mean in each direction is calculated to generate the dynamic fluctuation coefficient of the force ratio.

[0053] Evaluate the cutting process stability index based on the dynamic fluctuation coefficient of force ratio;

[0054] Based on the cutting process stability index, the thread profile cutting force characteristic data is correlated with the thread circle fluctuation amplitude to generate the cutting force dynamic fluctuation characteristic data.

[0055] In an embodiment of the present invention, a pitch preset value is obtained based on the fastener thread specification standard. For M10×1.5 specification fasteners, the standard pitch is 1.5 mm. The thread profile cutting force characteristic data is periodically segmented using a high-precision signal segmentation algorithm. The segmentation process first determines the starting point of thread processing, which is determined by the cutting force signal amplitude mutation point detection method, and the mutation threshold is set to 30% of the cutting force reference value. Starting from the starting point, the number of sampling points N within a single pitch is calculated based on the feed speed and sampling frequency of the CNC machine tool. The calculation formula is: N=f·p / v, where N represents the number of sampling points within a single pitch; f represents the sampling frequency, in Hertz, with a value of 10,000 Hertz; p represents the pitch value, in millimeters, with a value of 1.5 mm; v represents the feed speed, in millimeters / second, with a value of 2 mm / second. The entire cutting force data sequence is divided into m equally spaced segments according to the calculated number of periodic points N, resulting in m single-pitch cutting force periodic data segments. Each segment contains time series data for radial, axial, and tangential forces. A statistical average is performed on each single-pitch cutting force periodic data segment. The calculation method is to sum the radial force Fx_i(j), axial force Fy_i(j), and tangential force Fz_i(j) within the i-th periodic data segment and divide the sum by the number of sampling points N within that period, where j represents the sampling point number within the period, ranging from 1 to N. The calculation formula is: Fx_avg(i) = (1 / N)·∑Fx_i(j), Fy_avg(i) = (1 / N)·∑Fy_i(j), Fz_avg(i) = (1 / N)·∑Fz_i(j), where Fx_avg(i), Fy_avg(i), and Fz_avg(i) represent the average radial, axial, and tangential forces, respectively, within the i-th cycle, and are expressed in Newtons. Double-precision floating-point arithmetic is used to ensure accuracy to four decimal places. The three-dimensional average cutting forces of all cycles are combined into a three-dimensional vector: F_avg(i) = [Fx_avg(i), Fy_avg(i), Fz_avg(i)]. This generates the cycle-average cutting force vector data in the form of an m × 3 matrix, where m is the total number of thread cycles. A sliding window maximum detection algorithm is used to extract peak values ​​from the single-pitch cutting force cycle data. The window width is set to 5% of the number of cycle points, and the sliding step is 20% of the window width. For each cutting force component, within the i-th cycle, a sampling point is found that satisfies the local maximum condition. The local maximum condition is defined as: the value of the current point is greater than the values ​​of the adjacent points before and after it within the window, and greater than 1.2 times the average value within the window. The maximum points that meet these conditions are sorted, and the three largest peak points are selected. The average of these three peaks is calculated as the characteristic peak value of this component in that cycle.The calculation formulas are: Fx_peak(i) = (1 / 3)·∑Fx_i(pk), Fy_peak(i) = (1 / 3)·∑Fy_i(pk), Fz_peak(i) = (1 / 3)·∑Fz_i(pk), where pk represents the index of the kth peak point, and k is 1, 2, or 3. Fx_peak(i), Fy_peak(i), and Fz_peak(i) represent the characteristic peak values ​​of the radial force, axial force, and tangential force, respectively, within the i-th cycle, and are expressed in Newtons. Dynamic peak envelope processing is performed on the cycle-peak cutting force vector data using a cubic spline interpolation algorithm. The control points are the peak points of each cycle, and the boundary conditions are natural boundary conditions, where the second-order derivative is zero at the boundary. The number of interpolation points is five times the number of original cycles, generating a smooth peak envelope curve P(t). Similarly, the cycle-averaged cutting force vector data is processed using a dynamic mean trajectory to generate a smooth mean trajectory curve A(t). The ratios of the instantaneous cutting force peak to mean in the radial, axial, and tangential directions are calculated using the following formulas: Rx(t) = Px(t) / Ax(t), Ry(t) = Py(t) / Ay(t), and Rz(t) = Pz(t) / Az(t), where Rx(t), Ry(t), and Rz(t) represent the radial, axial, and tangential force ratios, respectively; Px(t), Py(t), and Pz(t) represent the components of the peak envelope curve in the three directions; and Ax(t), Ay(t), and Az(t) represent the components of the mean trajectory curve in the three directions. The dynamic fluctuation coefficient of the force ratio is obtained by synthesizing the three-axis force ratios. The cutting process stability is evaluated based on the statistical characteristics of the dynamic fluctuation coefficient of force ratio R(t). The mean μR, standard deviation σR, coefficient of variation CVR, and ratio of maximum value to mean RmaxR of R(t) are calculated. The calculation formula is: μR = (1 / L) · ∑R(tl), CVR = σR / μR, RmaxR = max(R(t)) / μR, where L represents the total number of sampling points; tl represents the lth sampling time; μR represents the dimensionless mean of the dynamic fluctuation coefficient of the force ratio; σR represents the dimensionless standard deviation; CVR represents the dimensionless coefficient of variation; and RmaxR represents the dimensionless ratio of the maximum value to the mean. The stability index S is constructed based on the coefficient of variation CVR and the maximum mean ratio RmaxR. The calculation formula is: S = 100 - 50·CVR-10·(RmaxR-1), where S represents the stability index, ranging from 0 to 100. When S is greater than 85, it is considered high stability; when S is between 65 and 85, it is considered moderate stability; and when S is less than 65, it is considered low stability. The cutting process stability index S is combined with the dynamic fluctuation coefficient of the force ratio R(t) to analyze the correlation between the fluctuation amplitude of the thread turns. First, the entire thread is divided into continuous thread turns, each of which contains periodic data for a standard number of thread pitches. For M10×1.5 fasteners, each standard thread contains 6.67 pitches, rounded to 7 pitches. The statistical characteristics of the dynamic fluctuation coefficient of the force ratio within each thread turn were calculated, including the mean, peak value, fluctuation range, and fluctuation frequency. The fluctuation frequency was calculated using a fast Fourier transform (FFT) with a sampling frequency of 10,000 Hz, 4,096 transformation points, and a frequency resolution of 2.44 Hz. A correlation model was established between the thread turn fluctuation amplitude and machining quality, and the correlation was calculated using the Pearson correlation coefficient method. When the absolute value of the correlation coefficient is greater than 0.8, it is considered a strong correlation; between 0.5 and 0.8, it is considered a moderate correlation; and less than 0.5, it is considered a weak correlation. Dynamic fluctuation characteristic data of cutting forces are generated, including the fluctuation characteristic value, stability index, and correlation rating for each thread turn, forming a complete thread machining quality assessment report.

[0056] Preferably, step S2 includes the following steps:

[0057] Step S21: performing cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data, and performing fluctuation-tool wear sensitivity matching to generate a tool wear indicator factor;

[0058] Step S22: obtaining reference average cutting force vector data and reference cutting force peak-to-average ratio data based on the tool wear indicator factor;

[0059] Step S23: subtracting the corresponding components of the reference average cutting force vector data from the period average cutting force vector data to obtain average force deviation vector data;

[0060] Step S24: subtracting corresponding items from the reference cutting force peak-to-average ratio data using the force ratio dynamic fluctuation coefficient to obtain a peak-to-average ratio deviation sequence;

[0061] Step S25: Identify the real-time abnormal situation of the current machining process based on the average force deviation vector data and the peak-to-average ratio deviation sequence to obtain real-time abnormal machining situation data.

[0062] In an embodiment of the present invention, a sample entropy algorithm is adopted, the embedding dimension is set to 2, the similarity tolerance is 0.15 times the standard deviation, and the scale factor range is 1 to 20. The entropy value at each scale is calculated, and the entropy spectrum curve E(s) is constructed, where s represents the scale factor. The spectral characteristics of the cutting force waveform are extracted by Fourier transform, with a sampling frequency of 10,000 Hz, a transformation length of 4,096 points, and a frequency resolution of 2.44 Hz. The main frequency component and its amplitude are extracted to form a frequency-amplitude pair. The entropy spectrum curve E(s) is combined with the spectral characteristics to establish a comprehensive evaluation index V of the vibration characteristics. The calculation formula is: V = α·∑w(s)·E(s)+β·∑A(f)·B(f), where α and β are weight coefficients, with values ​​of 0.6 and 0.4 respectively; w(s) is the scale weight function; A(f) is the frequency amplitude; and B(f) is the frequency weight function. A mapping relationship was established between the vibration characteristic evaluation index V and tool wear W. The wear was obtained through microscopic observation of the tool with an accuracy of 0.001 mm. This mapping was performed using a support vector regression algorithm, with a radial basis function kernel, a penalty factor C set to 10, and an epsilon parameter set to 0.01. A tool wear indicator factor D was generated using the following expression: D = V·(1+γ·T), where γ is the time adjustment factor, which is 0.02 / minute, and T is the cumulative cutting time in minutes. Based on the generated tool wear indicator factor D, baseline data collection conditions were determined. When the D value was less than 0.1, the tool was in good condition, and baseline data was collected at this time. Ten standard fastener samples were machined using a new tool, and cutting force data was collected throughout the entire process. The sampling frequency was set to 10,000 Hz, and the data collection time for each sample was 30 seconds. The collected data was filtered using a Butterworth low-pass filter with a cutoff frequency of 1,000 Hz and a filter order of 4. The filtered data is calculated according to the method of step S1 to calculate the cycle average cutting force vector and the dynamic fluctuation coefficient of the force ratio. The data of the 10 samples are averaged to calculate the benchmark average cutting force vector data F_base_avg = [Fx_base_avg, Fy_base_avg, Fz_base_avg], where Fx_base_avg, Fy_base_avg, and Fz_base_avg represent the benchmark average cutting forces in the radial, axial, and tangential directions, respectively, in Newtons. At the same time, the benchmark cutting force peak-to-average ratio data R_base = [Rx_base, Ry_base, Rz_base, R_total_base] is calculated, where Rx_base, Ry_base, and Rz_base represent the benchmark peak-to-average ratios in the radial, axial, and tangential directions, respectively, and R_total_base represents the benchmark peak-to-average ratio of the combined force, all of which are dimensionless. The corresponding components of the cycle average cutting force vector data obtained in real time are subtracted from the benchmark average cutting force vector data.The calculation process uses vector subtraction operation, and the formula is: ΔF_avg(i) = F_avg(i) - F_base_avg, that is, ΔFx_avg(i) = Fx_avg(i) - Fx_base_avg, ΔFy_avg(i) = Fy_avg(i) - Fy_base_avg, ΔFz_avg(i) = Fz_avg(i) - Fz_base_avg, where ΔFx_avg(i), ΔFy_avg(i), and ΔFz_avg(i) represent the average deviation of radial force, axial force, and tangential force in the i-th cycle, respectively, and the unit is Newton. To eliminate the impact of cutting parameter fluctuations, normalization is introduced and relative deviations are calculated using the following formulas: δFx_avg(i) = ΔFx_avg(i) / Fx_base_avg, δFy_avg(i) = ΔFy_avg(i) / Fy_base_avg, and δFz_avg(i) = ΔFz_avg(i) / Fz_base_avg, where δFx_avg(i), δFy_avg(i), and δFz_avg(i) represent the relative average deviations of radial force, axial force, and tangential force, respectively, and are dimensionless. The relative deviations in the three directions are combined into a vector form: δF_avg(i) = [δFx_avg(i), δFy_avg(i), δFz_avg(i)], yielding the average force deviation vector data. The real-time calculated force ratio dynamic fluctuation coefficient R(t) = [Rx(t), Ry(t), Rz(t), R_total(t)] is subtracted from the benchmark cutting force peak-to-average ratio data R_base = [Rx_base, Ry_base, Rz_base, R_total_base]. The subtraction operation uses the time series difference calculation method. For each time t, the calculation formulas are: ΔRx(t) = Rx(t) - Rx_base, ΔRy(t) = Ry(t) - Ry_base, ΔRz(t) = Rz(t) - Rz_base, ΔR_total(t) = R_total(t) - R_total_base, where ΔRx(t), ΔRy(t), and ΔRz(t) represent the peak-to-average ratio deviations in the radial, axial, and tangential directions, respectively, and ΔR_total(t) represents the peak-to-average ratio deviation of the resultant force. All are dimensionless. In order to eliminate the influence of the magnitude difference of the peak-to-average ratio in different directions, the relative deviation calculation is introduced, and the formula is: δRx(t) = ΔRx(t) / Rx_base, δRy(t) = ΔRy(t) / Ry_base, δRz(t) = ΔRz(t) / Rz_base.

[0063] δR_total(t)=ΔR_total(t) / R_total_base, where δRx(t), δRy(t), δRz(t), and δR_total(t) represent the relative peak-to-average deviations of each direction and the combined force, respectively, and are dimensionless. The relative peak-to-average deviations at all moments are arranged in chronological order to form a peak-to-average deviation sequence δR={δR(t1), δR(t2), ..., δR(tL)}, where L represents the total number of sampling points. Based on the average force deviation vector data δF_avg(i) and the peak-to-average deviation sequence δR, a two-dimensional feature space is constructed for abnormal situation recognition. First, the modulus of the average force deviation vector is calculated. As a feature dimension. Calculate the RMS value of the peak-to-average ratio deviation sequence As another feature dimension. The normal area and the abnormal area are divided in the two-dimensional feature space. The normal area is defined as: ‖δF_avg(i)‖<0.15 and RMSR<0.12. Points beyond the boundary of the normal area are judged as abnormal points. Cluster analysis is performed on the abnormal points, and the density clustering algorithm DBSCAN is used. The neighborhood radius ε is set to 0.05 and the minimum sample number MinPts is set to 5. According to the clustering results, the abnormal situation is divided into three categories: tool wear abnormality, cutting parameter abnormality and material property abnormality. The specific abnormality type is determined by the discriminant function H, which is constructed based on the directional characteristics of δF_avg(i) and δR and the time series pattern characteristics. Real-time abnormal processing situation data is generated, including abnormality type, abnormality degree, abnormality location and abnormality trend prediction. The data update frequency is 1 Hz, realizing real-time abnormality monitoring of the fastener thread processing process.

[0064] Preferably, scanning the local area profile of the machined thread surface of the fastener and calculating the local thread deviation value in step S3 includes:

[0065] Scan the contour of the local area of ​​the processed thread surface of the fastener to generate the local 3D topography point cloud data of the thread;

[0066] The fastener's spatial position is synchronized based on the local 3D thread topography point cloud data. Local fitting is then performed on the instantaneous distance change of continuous tooth tops or roots along the axial direction, the radial position of the middle point of the tooth profile, the instantaneous slope of the tooth profile side profile, and the tooth top arc shape to obtain the instantaneous thread geometric characteristic parameter data. Among them, the instantaneous thread geometric characteristic parameter data includes the instantaneous expansion and contraction of the pitch, the instantaneous radial offset of the thread effective diameter, the instantaneous distortion of the thread profile angle, and the instantaneous curvature change of the thread top fillet radius.

[0067] Based on the instantaneous geometric characteristic parameter data of the thread and the preset target geometric characteristic value of the thread, a real-time deviation calculation is performed to obtain the instantaneous geometric deviation data of the thread;

[0068] When the instantaneous expansion and contraction of the thread pitch in the instantaneous geometric deviation data shows a negative cumulative trend for three consecutive pitches, it is determined to be in the state of cumulative pitch contraction;

[0069] When the instantaneous distortion of the thread profile angle in the thread instantaneous geometric deviation data continuously exceeds the preset angle deviation value threshold of 0.5 degrees, it is determined to be a thread profile angle distortion state;

[0070] When the instantaneous radial deviation of the effective diameter of the thread in the instantaneous geometric deviation data exceeds 0.005 mm continuously, the effective diameter deviation state is obtained;

[0071] When the instantaneous curvature change of the thread top fillet radius in the thread instantaneous geometric deviation data continuously exceeds the preset relative reference curvature change by 5%, the thread top fillet curvature deviation state is obtained;

[0072] Dynamically assign weights to the instantaneous geometric deviation data of the thread to obtain the thread local deviation weight data;

[0073] The thread local deviation weight data is used to perform weighted aggregation calculation on the cumulative state of pitch shrinkage, tooth profile angle distortion, effective diameter offset and top fillet curvature deviation to generate the thread local accuracy deviation value.

[0074] In an embodiment of the present invention, an ultra-depth-of-field three-dimensional optical profiler is used to scan the machined threaded surface of a fastener. The profiler model is Keyence VR-6000, with a vertical resolution of 0.1 micron and a horizontal resolution of 0.5 micron. The scanning range is set to 10 mm along the thread axis and 120 degrees circumferentially. The scanning step size is set to 0.2 micron, the scanning speed is 10 mm / s, the light source intensity is 75%, and the exposure time is 20 milliseconds. During the scanning process, the fastener is fixed by a precision fixture with a positioning accuracy of ±0.002 mm and a repeatability accuracy of ±0.001 mm. The density of the collected raw point cloud data reaches 25,000 points per square millimeter, and the total data volume is approximately 15 million points. The raw point cloud data is denoised using a statistical outlier filtering algorithm with a neighborhood radius set to 0.05 mm and a standard deviation multiple threshold set to 2.5. The filtered point cloud data is downsampled and the voxel grid method is used with a grid size of 0.01 mm to generate regular thread local three-dimensional morphology point cloud data. The point cloud data is stored in the form of three-dimensional coordinates (x, y, z) with an accuracy of 6 decimal places. The main axis extraction algorithm is used to determine the thread axis. The algorithm uses the RANSAC cylindrical fitting method, the number of iterations is set to 1000 times, and the fitting threshold is 0.01 mm. A cylindrical coordinate system with the thread axis as the Z axis is established with a resolution of 0.001 mm. The tooth vertex set is extracted. The extraction condition is the point set with the largest distance from the thread axis. The number of points extracted is 30 points for each pitch. The tooth bottom point set is extracted. The extraction condition is the point set with the smallest distance from the thread axis. The number of points extracted is 25 points for each pitch. For a continuous set of tooth vertices, the axial distance between adjacent tooth crests is calculated to obtain the instantaneous pitch expansion and contraction ΔP(i). The calculation formula is: ΔP(i) = P(i) - P_std, where ΔP(i) represents the expansion and contraction of the i-th pitch, in millimeters; P(i) represents the actual measured i-th pitch, in millimeters; and P_std represents the standard pitch value, which is 1.5 mm for M10×1.5 fasteners. The midpoint of the tooth profile is extracted, located at the midpoint of the line connecting the tooth crest and root, and its radial position, r_mid(i), is calculated to obtain the instantaneous radial deviation of the effective thread diameter, Δr(i) = r_mid(i) - r_std, where r_std is the standard effective radius, which is 4.513 mm. The point set on the side of the tooth profile is plane fitted using the least squares method. The fitting accuracy requires the root mean square error to be less than 0.002 mm. The angle θ(i) between the fitting plane and the radial plane is calculated to obtain the instantaneous distortion of the thread profile angle Δθ(i) = θ(i) - θ_std, where θ_std is the standard tooth profile angle, which is 30 degrees.Arc fitting is performed on the tooth vertex set using an iterative closest point algorithm with 100 iterations and a convergence threshold of 0.0001 mm. The tip fillet radius R(i) is then calculated, and the instantaneous curvature change of the thread tip fillet radius ΔK(i) = 1 / R(i) - 1 / R_std, where R_std is the standard tip fillet radius, which is 0.144 mm. The preset target thread geometric characteristic values ​​are obtained from the national standard GB / T193-2003, "Tolerances for General Threads." For M10×1.5 fasteners, the 6g tolerance grade has an allowable pitch deviation of ±0.045 mm, an effective diameter deviation of -0.022 to -0.104 mm, a thread profile angle deviation of ±0.75 degrees, and a tip fillet radius deviation of ±15%. Calculate the relative deviation of the instantaneous expansion and contraction of the thread pitch: εP(i) = ΔP(i) / P_std, where εP(i) represents the relative expansion and contraction of the i-th pitch and is dimensionless. Calculate the relative deviation of the instantaneous radial offset of the effective diameter of the thread: εr(i) = Δr(i) / r_std, where εr(i) represents the relative radial offset of the effective diameter at the i-th position and is dimensionless. Calculate the relative deviation of the instantaneous distortion of the thread profile angle: εθ(i) = Δθ(i) / θ_std, where εθ(i) represents the relative distortion of the profile angle at the i-th position and is dimensionless. Calculate the relative deviation of the instantaneous curvature change of the thread tip fillet radius: εK(i) = ΔK(i) / (1 / R_std), where εK(i) represents the relative curvature change of the tip fillet at the i-th position and is dimensionless. All relative deviation data were combined into a dataset of instantaneous thread geometric deviations {εP(i), εr(i), εθ(i), εK(i)}. A three-point sliding window method was used to analyze the trend of three consecutive thread pitches, with the data within the window being {εP(i), εP(i+1), εP(i+2)}. The first-order difference sequence {ΔεP(i), ΔεP(i+1)} of the data within the window was calculated, where ΔεP(i) = εP(i+1) - εP(i) and ΔεP(i+1) = εP(i+2) - εP(i+1). When ΔεP(i) < 0 and ΔεP(i+1) < 0, it was considered a continuous negative change. The cumulative change Σ = εP(i) + εP(i+1) + εP(i+2) was also calculated. When Σ < -0.01, the cumulative effect was considered significant. When both the continuous negative change and the significant cumulative effect conditions are met, the current window is marked as the pitch contraction cumulative state. The window sliding step is 1, and the entire thread area is analyzed. For each window marked as the pitch contraction cumulative state, its starting position, length and cumulative change are recorded to generate a pitch contraction cumulative state data record. When the number of pitch contraction cumulative state data records exceeds 20% of the total number of thread pitches, a pitch contraction cumulative abnormality warning is issued. The angle offset numerical threshold is set to 0.5 degrees, and the corresponding relative deviation threshold εθ_threshold = 0.5 / 30 = 0.0167.The εθ(i) value at each position i is evaluated. When |εθ(i)| > εθ_threshold, the position is marked as an angular anomaly. A persistence judgment algorithm is used, defining the persistence interval length as five consecutive measurement points. When five or more consecutive measurement points are marked as angular anomalies, the interval is considered to be in a state of tooth profile angle distortion. The average deviation value within the distortion interval is calculated as μθ = (1 / n)·∑εθ(i), where n is the number of measurement points in the interval. The distortion direction is determined based on the sign of μθ: μθ > 0 indicates an increase in the angle, and μθ < 0 indicates a decrease in the angle. The starting position, ending position, length, average deviation value, and distortion direction of the distortion interval are recorded to generate a tooth profile angle distortion status data record. When the cumulative interval length of the tooth profile angle distortion state exceeds 15% of the total thread length, a tooth profile angle distortion anomaly warning is issued. The offset threshold is set to 0.005 mm, and the actual offset value Δr(i) at each position i is compared with the threshold. When |Δr(i)|>0.005 mm, the location is marked as a diameter outlier. Cluster analysis is used to identify continuous outlier regions. The clustering algorithm uses density clustering (DBSCAN) with parameters set to ε = 0.3 mm (spatial proximity) and MinPts = 5 (minimum number of points). Each formed cluster represents a continuous outlier region, referred to as an effective diameter deviation state interval. The average deviation value within each interval is calculated as μr = (1 / m)·∑Δr(j), where m is the number of measurement points within the interval and j is the index of the point within the interval. The deviation direction is determined based on the sign of μr: μr > 0 indicates an increase in diameter, and μr < 0 indicates a decrease in diameter. The spatial location, length, average deviation value, and deviation direction of each effective diameter deviation state interval are recorded to generate effective diameter deviation state data records. When the cumulative interval coverage angle of the effective diameter deviation state exceeds 30% of the total circumferential angle of the thread, an effective diameter deviation anomaly warning is issued. The relative baseline curvature change threshold is set to 5%, i.e., εK_threshold = 0.05. The εK(i) value at each location i is evaluated. When |εK(i)| > εK_threshold, the location is marked as a curvature outlier. A region growing algorithm is used to identify continuous outlier regions. The seed point selection condition is the local maximum deviation point, and the growth condition is that the deviation values ​​of adjacent points are greater than the threshold and the deviation direction is consistent. The growth parameters are set to a neighborhood radius of 0.2 mm and a similarity threshold of 0.02. Each region after growth represents a crown fillet curvature deviation state interval. The average curvature deviation value within the interval is calculated as μK = (1 / p) · ∑εK(k), where p is the number of measurement points in the interval and k is the index of the point within the interval. The direction of curvature change is determined by the positive or negative value of μK: μK > 0 indicates an increase in curvature (decrease in radius), and μK < 0 indicates a decrease in curvature (increase in radius). The spatial location, coverage, average deviation value, and change direction of each crown fillet curvature deviation state interval are recorded to generate crown fillet curvature deviation state data records.When the cumulative interval length of the top fillet curvature deviation state exceeds 25% of the total thread length, a top fillet curvature deviation abnormality warning is issued. An importance analysis is performed on the four states (pitch shrinkage cumulative state, tooth profile angle distortion state, effective diameter offset state, and top fillet curvature deviation state) to establish a dynamic weight distribution mechanism. First, based on the degree of influence on the fastener assembly performance, the basic weight coefficients are set: pitch shrinkage cumulative state w_P_base = 0.35, tooth profile angle distortion state w_θ_base = 0.25, effective diameter offset state w_r_base = 0.30, and top fillet curvature deviation state w_K_base = 0.10. Then, the deviation severity adjustment factor is introduced, and the calculation method is: g_P = min(|∑p| / 0.03, 2), g_θ = min(|μθ| / 0.0167, 2), g_r = min(|μr| / 0.005, 2), g_K = min(|μK| / 0.05, 2), where g_P, g_θ, g_r, and g_K represent the severity adjustment factors of the four states, respectively, and are dimensionless. A coverage adjustment factor is introduced and calculated as follows: h_P = min(L_P / L_total, 1), h_θ = min(L_θ / L_total, 1), h_r = min(L_r / L_total, 1), and h_K = min(L_K / L_total, 1). h_P, h_θ, h_r, and h_K are dimensionless coverage adjustment factors for the four states. L_P, L_θ, L_r, and L_K represent the cumulative abnormal interval lengths for the four states, and L_total represents the total thread length. The degree of deviation for each state is normalized and converted to a score between 0 and 1. The cumulative score for pitch shrinkage, S_P, is min(|Σ| / 0.03, 1), the score for tooth profile angle distortion, S_θ, is min(|μθ| / 0.0167, 1), the score for effective diameter offset, S_r, is min(|μr| / 0.005, 1), and the score for top fillet curvature deviation, S_K, is min(|μK| / 0.05, 1). The weighted average method is then used to calculate the thread local accuracy deviation, D, using the formula: D = (w_P·S_P+w_θ·S_θ+w_r·S_r+w_K·S_K) / (w_P+w_θ+w_r+w_K)·100, where D represents the thread local accuracy deviation and ranges from 0 to 100. A spatial position weighting coefficient is introduced to increase the influence weight of abnormal areas in key thread locations (such as the first and last three threads) by 50%. The final precision deviation value is accurate to 2 decimal places. When the D value is less than 20, it is judged as a high-precision thread; when the D value is between 20 and 50, it is judged as a standard-precision thread; when the D value is between 50 and 80, it is judged as a low-precision thread; when the D value is greater than 80, it is judged as an unqualified thread.Generate a thread local accuracy deviation assessment report, including the overall deviation value, details of each deviation status, and spatial distribution diagram.

[0075] It is particularly important to dynamically allocate weights to the instantaneous thread geometry deviation data as follows:

[0076] The initial evaluation data of the pitch error is obtained by comparing the instantaneous expansion and contraction of the pitch in the instantaneous geometric deviation data with the preset pitch cumulative error threshold;

[0077] Comparing the instantaneous radial offset of the thread effective diameter in the instantaneous geometric deviation data with a preset effective diameter deviation threshold value to obtain initial evaluation data of the effective diameter deviation;

[0078] The initial evaluation data of the thread profile angle distortion is obtained by comparing the instantaneous distortion of the thread profile angle in the instantaneous thread geometric deviation data with the preset thread profile angle distortion threshold;

[0079] The initial evaluation data of the top fillet radius is obtained by comparing the instantaneous curvature change of the thread top fillet radius in the thread instantaneous geometric deviation data with the preset instantaneous curvature change threshold of the thread top fillet radius;

[0080] Dynamic weight allocation is performed on the initial evaluation data of pitch error, effective diameter deviation, tooth profile angle distortion and top fillet radius to obtain thread local deviation weight data.

[0081] In this embodiment of the present invention, for M10×1.5 fasteners, the 6g tolerance grade pitch cumulative error threshold is set to ±0.045 mm. This threshold is calibrated using a high-precision pitch micrometer, and the calibration repeatability error is controlled within ±0.001 mm. The instantaneous pitch expansion and contraction εP(i) in the thread instantaneous geometric deviation data is compared with a preset threshold, and the error severity index SP(i) is calculated as |εP(i)| / 0.03, where 0.03 is the standardized reference value, taken from 2 / 3 of the tolerance band. When the εP(i) values ​​of three consecutive pitches are accumulated in the same direction, the cumulative value AC(i) is calculated as |εP(i)+εP(i+1)+εP(i+2), and the cumulative coefficient FC(i) is calculated as |AC(i)| / 0.045. Spatial position weighting is introduced, and the position coefficient LP(i) of the first three and last three threads of the fastener is set to 1.5, and that of the middle region is set to 1.0. The comprehensive error score EP(i) = SP(i)·FC(i)·LP(i) is calculated, ranging from 0 to 10. The entire thread area is divided into five equal-length segments. The maximum EP value in each segment is selected to form the initial pitch error assessment data {EP_1, EP_2, EP_3, EP_4, EP_5}, with data accurate to two decimal places. The effective diameter deviation threshold is determined according to the national standard GB / T196-2003 "General Thread Tolerances." For M10×1.5 fasteners, the upper and lower effective diameter deviations for the 6g tolerance grade are -0.022 mm and -0.104 mm, respectively, with an effective range width of 0.082 mm. The instantaneous radial deviation of the effective diameter εr(i) in the thread instantaneous geometric deviation data is converted to the actual dimensional deviation Dr(i) = εr(i)·9.026, where 9.026 mm is the standard effective diameter of an M10×1.5 fastener. The deviation position coefficient PD(i) is calculated as (Dr(i) - (-0.022)) / 0.082. When PD(i) is less than 0 or greater than 1, it indicates an out-of-tolerance zone. The deviation severity index SD(i) is introduced and calculated as follows: when PD(i) < 0, SD(i) = 1 + |PD(i)|; when 0 ≤ PD(i) ≤ 1, SD(i) = 0; when PD(i) > 1, SD(i) = PD(i). Considering the circumferential distribution of measurement positions, four points are evenly selected on the thread circumference for measurement, and the circumferential consistency index CD is calculated as 1-min(σD / 0.01, 1), where σD is the standard deviation of the four circumferential measurement points. The effective diameter score ED(i) is calculated as SD(i) · (2-CD), ranging from 0 to 10. The thread area is divided into three regions: the front section, the middle section, and the back section. The maximum ED value in each region is selected to form the initial evaluation data of the effective diameter deviation. For metric threads, the standard tooth profile angle is 60 degrees, and the tooth profile angle distortion threshold is set to ±0.5 degrees, which is approximately equivalent to a relative deviation of ±0.0083.The instantaneous distortion of the thread profile angle εθ(i) in the instantaneous thread geometric deviation data is compared with the preset threshold. The angle deviation severity index SA(i) is calculated as |εθ(i)| / 0.0083, which represents the ratio of the actual deviation to the threshold. A persistence coefficient is introduced. When the εθ(i) values ​​of five consecutive measurement points deviate in the same direction and |εθ(i)|>0.0083, the persistence coefficient DF(i) is set to 1.5, otherwise it is set to 1.0. A multi-scale morphology analysis algorithm is used to evaluate the integrity of the thread profile and extract the contour line smoothness index LS(i). The calculation method is: Where RMS stands for root mean square, represents the second-order differential operator, and 0.001 is the normalization factor. The tooth profile angle score EA(i) = SA(i)·DF(i)·LS(i) is calculated comprehensively, with a score range of 0 to 10. The thread area is evenly divided into 6 segments along the thread axis, and the maximum EA value is selected in each segment to form the initial evaluation data of the tooth profile angle distortion. For M10×1.5 specification fasteners, the standard tooth top fillet radius is 0.144 mm, corresponding to a curvature of 6.94 / mm. The tooth top fillet radius curvature change threshold is set to ±5%, or ±0.347 / mm. The instantaneous curvature change εK(i) of the thread tooth top fillet radius in the thread instantaneous geometric deviation data is compared with the preset threshold. The curvature deviation severity index SK(i) = |εK(i)| / 0.05 is calculated, which represents the ratio of the actual relative deviation to the threshold. The influencing factor of the tooth top surface roughness is introduced. The surface roughness Ra(i) of the tooth top area is calculated using the three-dimensional topography data. The standard Ra value of the tooth top area is 0.8 microns. The surface state coefficient RS(i) is calculated as 1+max(Ra(i) / 0.8-1,0). Considering the continuity of the fillet, the curvature change rate evaluation index is used. in Represents the spatial gradient operator, and 0.01 is the normalization factor. Comprehensively calculate the top fillet score EK(i) = SK(i)·RS(i)·CG(i), with a score range of 0 to 10. Divide the thread into 8 sectors along the circumference, and select the maximum EK value in each sector to form the initial evaluation data of the top fillet radius. Calculate the average value of each initial evaluation data: average pitch error score MP = (EP_1+EP_2+EP_3+EP_4+EP_5) / 5; average effective diameter deviation score MD = (ED_1+ED_2+ED_3) / 3;

[0082] The average tooth profile angle distortion score (MA) is calculated as (EA_1+EA_2+EA_3+EA_4+EA_5+EA_6) / 6; the average tooth tip fillet radius score (MK) is calculated as (EK_1+EK_2+EK_3+EK_4+EK_5+EK_6+EK_7+EK_8) / 8. The following basic weights are set based on the fastener's functional requirements: the basic weight for pitch error (WP_base) is calculated as 0.30; the basic weight for effective diameter deviation (WD_base) is calculated as 0.35; the basic weight for tooth profile angle distortion (WA_base) is calculated as 0.25; and the basic weight for tip fillet radius (WK_base) is calculated as 0.10. A dynamic adjustment factor is introduced, and the calculation formula is: TP = min(MP / 5, 1)·0.5+0.5; TD = min(MD / 5, 1)·0.5+0.5; TA = min(MA / 5, 1)·0.5+0.5; TK = min(MK / 5, 1)·0.5+0.5.

[0083] Preferably, in step S3, judging the tool wear mode based on the real-time abnormal machining situation data and the thread local precision deviation value, and predicting the thread circle machining deviation includes:

[0084] The tool wear pattern is judged based on the local thread accuracy deviation value and real-time abnormal processing status data to generate tool wear pattern recognition data. The tool wear pattern judgment includes the tool tip / side edge grinding circular wear pattern and the main cutting edge local chipping wear pattern;

[0085] The wear degree of the tool wear pattern recognition data is quantified by the local thread accuracy deviation value, and the wear value of the tool tip radius and the wear degree of the side edge are estimated respectively.

[0086] Obtain the maximum tool tip radius wear and maximum side edge wear allowed by tool design;

[0087] According to the maximum tool tip radius wear and the maximum side edge wear, the tool tip radius wear estimation and side edge wear estimation are calculated in percentage, and the larger value of the two is taken as the comprehensive tool wear percentage;

[0088] Based on the tool comprehensive wear percentage, the tool wear pattern recognition data is used to predict the thread circle diameter deviation, pitch deviation and tooth profile angle deviation to obtain the predicted thread geometry deviation.

[0089] In an embodiment of the present invention, pattern recognition is performed based on a pre-established wear feature database. The database contains 2,000 sets of typical wear cases, each set of cases containing cutting force features, geometric deviation features, and corresponding wear pattern markers. For the tip / side grinding circular wear pattern, the performance is as follows: the cutting force peak-to-average ratio is lower than 1.35, the force fluctuation frequency is lower than 50 Hz, the effective diameter of the thread shows a uniform shrinking trend (deviation is greater than 0.02 mm), and the radius of the tooth top fillet increases (the curvature decreases by more than 7%). For the local chipping wear pattern of the main cutting edge, the performance is as follows: the cutting force mutation amplitude is greater than 30%, the force fluctuation frequency is higher than 80 Hz, the thread tooth profile angle is locally distorted (the angle deviation is greater than 0.8 degrees), and the tooth profile profile shows irregular jagged changes. The Mahalanobis distance is used to calculate the matching degree between the current data and each pattern in the feature library, and the distance threshold is set to 2.5. When the matching degree exceeds 85%, it is determined to be the corresponding wear pattern. When two patterns exist at the same time, the dominant pattern is determined by principal component analysis, and the weight ratio is calculated according to the feature contribution. Tool wear pattern recognition data is generated, including wear pattern type, matching degree, main characteristic parameters, and confidence score. For tool tip radius wear estimation, the corresponding relationship between the curvature change of the thread top fillet radius and tool tip radius wear is used for calculation. A mapping function is established: Tool tip radius wear estimation RT = 0.05 + 0.8 · |εK_avg| · (1 + 0.2 · D), where RT represents the tool tip radius wear estimation in millimeters; εK_avg represents the dimensionless average curvature change of the thread top fillet radius; and D represents the thread local accuracy deviation, ranging from 0 to 100. For side edge wear estimation, the corresponding relationship between thread effective diameter offset and side edge wear is used for calculation. A mapping function is established: Side edge wear estimation RS = 0.03 + 0.6 · |εr_avg| · (1 + 0.15 · D), where RS represents the side edge wear estimation in millimeters; and εr_avg represents the dimensionless average relative offset of the thread effective diameter. During the estimation process, correction factors are used to correct the machining conditions. These correction factors take into account the workpiece material hardness (based on HB240), cutting speed (based on 60 m / min), and cutting depth (based on 0.08 mm). The correction factors are calculated using a wear accumulation effect model. For every 50 HB increase in workpiece hardness, the wear estimate increases by 15%; for every 20 m / min increase in cutting speed, the wear estimate increases by 12%; and for every 0.02 mm increase in cutting depth, the wear estimate increases by 8%. The maximum wear allowable for tool design is obtained from the tool technical parameter database. This database contains detailed technical parameters for different tool models and is stored in the production control host. The data is updated monthly. For the standard turning tool used for threading M10×1.5 fasteners, model DCMT11T304-PM, the maximum tool tip radius wear is 0.25 mm and the maximum side edge wear is 0.18 mm.These data are based on the technical specifications provided by the tool manufacturer and were determined after experimental verification. The verification experiment utilized electron microscopy observation using a Nikon MM-800 measuring microscope with a magnification of 200x and a measurement accuracy of 0.001 mm. During the experiment, changes in cutting force and machining quality under different wear conditions were recorded to determine the critical wear value. Considering a process safety factor of 1.2, the theoretical critical value was divided by the safety factor to obtain the actual maximum allowable wear. The maximum allowable wear varies for different tool materials: the maximum tip radius wear for carbide tools (YG8) is 0.25 mm, for high-speed steel tools it is 0.20 mm, and for ceramic tools it is 0.18 mm. The formula for calculating the tip radius wear percentage is: PT = (RT / RT_max)·100%, where PT represents the tip radius wear percentage; RT represents the estimated tip radius wear value in millimeters; and RT_max represents the maximum tip radius wear in millimeters. The formula for calculating the swarf wear percentage is: PS = (RS / RS_max)·100%, where PS represents the swarf wear percentage; RS represents the estimated swarf wear level (in millimeters); and RS_max represents the maximum swarf wear (in millimeters). A comparison algorithm is used to determine the larger of the two values ​​as the tool's overall wear percentage, P, calculated as: P = max(PT, PS). To improve calculation accuracy, a machining condition correction factor is introduced to account for the cumulative effects of cutting time, cutting temperature, and cooling conditions. The cutting time cumulative effect coefficient, KT, is proportional to the cumulative cutting time, T, as KT = 1 + 0.005·T, where T is measured in minutes. The cutting temperature influence coefficient, KΘ, is proportional to the cutting zone temperature, Θ, as KΘ = 1 + 0.002·(Θ - 300°C), where Θ is measured in degrees Celsius using an infrared thermometer. The cooling condition influence coefficient, KC, is determined based on the coolant flow rate, F, as KC = 1.1 - 0.005·F, where F is measured in liters per minute. The final comprehensive wear percentage is: P′ = P·KT·KΘ·KC. A prediction model based on a neural network was constructed. The network structure is a three-layer feedforward network with 8 neurons in the input layer, 12 neurons in the hidden layer, and 3 neurons in the output layer. The input parameters include: the comprehensive tool wear percentage P′, the wear pattern type (binary encoding), the current thread local accuracy deviation value D, the cutting speed v, the feed rate f, the cutting depth ap, the workpiece material hardness HB, and the cooling condition index C. The output parameters are the predicted geometric deviations: the thread circle diameter deviation ΔD, the pitch deviation ΔP, and the tooth profile angle deviation Δθ. The network training uses 2000 sets of historical processing data and the Levenberg-Marquardt algorithm. The training accuracy is set to 0.001 and the maximum number of iterations is 5000.For the tool tip / side grinding circular wear pattern, the prediction model focuses on diameter deviation and tip radius variation. For the main cutting edge local chipping wear pattern, the prediction model focuses on tooth profile angle deviation and pitch variation. The model's prediction accuracy was evaluated through residual analysis, with average prediction errors within ±0.005 mm for diameter deviation, ±0.008 mm for pitch deviation, and ±0.15 degrees for angle deviation. The prediction output is the predicted geometric deviation data for the next thread revolution, including the predicted value, confidence interval, and risk level assessment.

[0090] Preferably, in step S4, adjusting multiple processing control parameter values ​​according to the predicted thread geometry deviation includes:

[0091] The compensation value of the spindle speed and feed axis synchronization ratio is calculated based on the pitch deviation in the predicted thread geometric deviation, and the instantaneous phase compensation instruction of the spindle is obtained;

[0092] Calculate the adjustment value of the radial cutting depth according to the thread circle diameter deviation in the predicted thread geometric deviation to obtain the cutting depth adjustment instruction;

[0093] Based on the real-time abnormal machining situation data, the adjustment value of the cooling parameter is calculated based on the tooth profile angle offset in the predicted thread geometry deviation, and the cooling feed adjustment instruction is obtained;

[0094] The spindle instantaneous phase compensation instruction, cutting depth adjustment instruction and cooling feed adjustment instruction are combined into processing control parameters to obtain the compound thread processing adjustment instruction.

[0095] In this embodiment of the present invention, the standard pitch P_std is determined. For M10×1.5 fasteners, P_std is 1.5 mm. The relative pitch deviation rate ε_p is calculated as ΔP / P_std. For a predicted deviation ΔP of 0.012 mm, the relative deviation rate is 0.8%. The relative deviation rate is converted into a spindle encoder pulse phase compensation value using the internal synchronization ratio adjustment algorithm of the CNC system. The CNC system is a FANUC 0i-TF, with a spindle encoder resolution of 10,000 pulses / rev and a feed axis encoder resolution of 5,000 pulses / mm. The compensation value is calculated using pulse interpolation technology, with the number of interpolated pulses per pitch being N_comp = 10,000 × ε_p = 80 pulses. The system performs instantaneous spindle phase fine-tuning with an adjustment range of ±150 pulses, an adjustment accuracy of 1 pulse, and an adjustment frequency of 200 Hz. Phase compensation achieves dynamic offset of the axis coordinate system using the G92.1 command, and the offset value is updated in real time based on the machining progress. When the pitch compensation value exceeds ±0.03 mm, the system issues a warning and records an exception log. A spindle instantaneous phase compensation command is generated in standard CNC G-code format, including the axis offset, execution time, and compensation duration. For M10×1.5 fasteners, radial cutting depth compensation is required when the predicted diameter deviation ΔD = -0.018 mm. In CNC thread machining, radial feed is typically performed using multiple feeds, with the standard value for the last cutting feed being 0.05 mm. The calculated radial compensation amount Δd = -ΔD / 2 = 0.009 mm, indicating that the radial cutting depth needs to be increased by 0.009 mm. Compensation is implemented using the CNC system's workpiece coordinate system fine-tuning function, with the G10L20 command modifying the X-axis workpiece coordinate system offset value. Compensation is implemented in a segmented incremental manner, starting at the starting position of the next thread turn and linearly increasing to the target compensation value within the first three pitches to avoid sudden changes in tool load. Adjustment accuracy is controlled within 0.001 mm, and servo driver parameter P5023 (position loop gain) is adjusted to 1.2 times the original value to improve positioning accuracy. Parameter P5081 (backlash compensation value) is also modified to ensure positioning accuracy during reciprocating motion. The actual cutting depth after compensation is monitored in real time by the CNC system, and the actual execution value is recorded through macro program #3101 to generate a cutting depth adjustment instruction. Cooling parameter adjustment values ​​are calculated based on the predicted thread profile angle offset Δθ in the thread geometry deviation based on real-time abnormal machining status data. When the predicted thread profile angle offset Δθ exceeds 0.3 degrees, it is determined to be thermal deformation caused by abnormal cutting temperature. The cooling system is a high-pressure precision cooling device, model KNOLLKF80-175, with a standard coolant pressure of 2.5 MPa, a flow rate of 15 liters / minute, and a nozzle diameter of 1.2 mm. The coolant pressure adjustment value ΔP_cool is calculated based on the thread profile angle offset, with a pressure adjustment of 0.2 MPa for every 0.1 degree of angle offset.For a predicted angular offset of Δθ = 0.4 degrees, the pressure adjustment value is 0.8 MPa, resulting in a pressure adjustment of 3.3 MPa. The coolant flow adjustment value, ΔF_cool, is also calculated, using a flow adjustment of 1.5 liters / minute for every 0.1-degree angular offset. The adjusted flow rate is 21 liters / minute. The cooling nozzle position is adjusted based on the tool wear pattern. For tool tip rounding, the nozzle angle is adjusted to 25 degrees toward the tool tip; for side edge chipping, the nozzle angle is adjusted to 40 degrees toward the side edge. The adjustment command is executed via PLC block DB100, which contains the pressure value, flow value, nozzle angle, start time, and duration. The spindle instantaneous phase compensation command, cutting depth adjustment command, and coolant feed adjustment command are combined to generate the composite thread machining adjustment command. This combination process is implemented using a CNC system macro program, numbered O9000, which utilizes parametric programming. The command combination sequence is as follows: cooling parameter adjustment is executed first (30 milliseconds in advance), followed by simultaneous execution of depth of cut adjustment and spindle phase compensation. The execution priority of each command is: spindle phase compensation (priority 1), depth of cut adjustment (priority 2), and cooling parameter adjustment (priority 3). In the event of a conflict between commands, the higher-priority command is executed. Time synchronization of command execution is achieved using a high-precision timer within the CNC system, with a timing accuracy of 0.1 millisecond. Compound commands include execution conditional logic. When the tool wear percentage exceeds 85%, the maximum spindle phase compensation value is limited to 80% of the original setting, and the maximum depth of cut adjustment value is limited to 70% of the original setting to protect the tool. Compound commands are transmitted to the execution unit via the CNC system's built-in Ethernet interface at a transmission rate of 100 Mbit / s, with transmission latency controlled to less than 1 millisecond. After receiving the command, the execution unit distributes it to each execution module via a real-time control bus (cycle time of 0.5 milliseconds), enabling coordinated adjustment of multiple parameters.

[0096] Preferably, the compensation value of the synchronization ratio between the spindle speed and the feed axis is calculated based on the pitch deviation in the predicted thread geometric deviation, including:

[0097] According to the pitch deviation in the predicted thread geometry deviation, the direction and value of the pitch cumulative error are identified, and the single-turn pitch compensation amount is evaluated to obtain the single-turn pitch compensation amount data;

[0098] Get the current spindle speed and thread lead data;

[0099] Based on the single-turn pitch compensation data, the instantaneous angular velocity adjustment amplitude that the spindle and the feed axis should perform in each thread spiral turn is calculated through the current spindle speed and thread lead data to obtain the spindle angular velocity adjustment amplitude data;

[0100] Taking every 30 degrees as an adjustment point, the spindle angular velocity adjustment amplitude data is decomposed into a sequence of instantaneous angular velocity adjustment amounts on the thread cutting path to generate the spindle instantaneous angular velocity adjustment data;

[0101] The instantaneous angular velocity of the main shaft is adjusted according to the data of the angular velocity modulation instruction, and the instantaneous phase compensation instruction of the main shaft is generated.

[0102] In the embodiment of the present invention, a linear trend extraction algorithm is used to process the pitch deviation data, and the trend of the pitch deviation changing with the axial position is fitted by the least squares method. The fitting accuracy requires that the residual square sum is less than 0.001 mm. 2. For M10×1.5 specification fasteners, when the predicted pitch deviation value is +0.015 mm, it means that the actual pitch is greater than the standard pitch and shows a loose trend; when the predicted deviation value is -0.012 mm, it means that the actual pitch is less than the standard pitch and shows a tightening trend. The cumulative error calculation adopts the step accumulation method, and the deviation values ​​of 7 consecutive pitches are accumulated to obtain the total cumulative error of a single turn. For standard M10×1.5 fasteners, one turn of thread contains 6.67 pitches, which is rounded to 7 pitches. The single-turn pitch compensation amount calculation adopts the piecewise linear interpolation method, and the opposite compensation amount is set according to the direction of the cumulative error within one spiral turn, with a compensation accuracy of 0.001 mm. The compensation amount is controlled between 80% and 120% of the predicted deviation to avoid oscillation caused by over-compensation. Generate single-turn pitch compensation amount data, the data format is an angle-compensation amount correspondence table, and the angular resolution is 10 degrees. The CNC system is a Siemens 840D. The actual spindle speed is read in rpm using the system variable $AA_S[S1], with an accuracy of 0.1 rpm. For machining M10×1.5 fasteners, the nominal spindle speed is set to 800 rpm. The actual thread lead is calculated using the feed axis parameters $AA_IM[X] and $AA_IM[Z] with an accuracy of 0.001 mm. Thread lead data includes the nominal lead value P_nom and the actual lead value P_real. For M10×1.5 fasteners, the nominal lead value is 1.5 mm. Lead deviation is measured using a high-precision measuring instrument, a Zeiss O-INSPECT, with an accuracy of 0.001 mm. The three-wire method is used, with a measuring force of 0.5 Newtons. Measurements are taken at the center of the thread at eight points evenly distributed around the circumference. When the spindle speed fluctuates by more than ±2% or the lead deviation exceeds ±0.008 mm, the system triggers a remeasurement procedure and updates the data in the parameter table. Data is collected at a 10 Hz frequency and transmitted to the control unit via a real-time bus. Based on the single-turn pitch compensation data, combined with the current spindle speed and thread lead data, the instantaneous angular velocity adjustment of the spindle and feed axis is calculated. First, the single-turn pitch compensation is converted to an angular domain compensation using the following formula: every 0.001 mm of pitch compensation corresponds to 0.24 degrees of angle compensation (calculated based on the helix angle of an M10×1.5 fastener). The rated spindle angular velocity ω_nom is calculated as 2π × spindle speed / 60, expressed in radians per second. For a spindle speed of 800 rpm, the rated angular velocity is 83.78 radians per second. The angular velocity adjustment ω_adj is calculated as ω_adj = single-turn compensation × 2π / (thread lead × spindle speed / 60), expressed in radians per second. The adjustment should be within ±5% of the rated angular velocity to avoid vibration caused by excessive spindle speed variations. The adjustment direction is opposite to the direction of the pitch deviation: when the pitch is too large, the spindle angular velocity needs to be increased; when the pitch is too small, the spindle angular velocity needs to be decreased.The calculated angular velocity adjustment amplitude is stored as discrete data with 36 data points per revolution, generating spindle angular velocity adjustment amplitude data. The spindle angular velocity adjustment amplitude data is then decomposed into a sequence of instantaneous angular velocity adjustments along the thread cutting path. A segmented processing method is used, with each adjustment point defined as 30 degrees, dividing a complete helical revolution into 12 adjustment segments. For each adjustment point, the instantaneous angular velocity adjustment required for that position is calculated. This adjustment calculation utilizes a cubic spline interpolation algorithm to ensure smooth and continuous angular velocity changes between adjacent adjustment points, avoiding mechanical shock caused by sudden changes. Interpolation points are evenly distributed, with 10 interpolation points per segment, for a total of 120 control points across the entire revolution. The angular velocity adjustment amplitude is smoothed using a cosine transition function at the start and end of the thread, with the transition zone length being 15% of the total thread length. The transition function is expressed as follows: the adjustment amplitude gradually increases from zero to its full value in the first segment and gradually decreases from its full value to zero in the second segment. Considering the dynamic response characteristics of the servo system, feedforward compensation is applied to the adjustment sequence, with the compensation time advanced by 1.5 milliseconds to overcome the servo system's phase lag. Spindle instantaneous angular velocity adjustment data is generated with a data point density of 10 points / degree. The CNC system's real-time interpolation module is used for command processing, directly controlling the spindle position loop. The angular velocity adjustment is converted to position increments using the following calculation: Position increment = angular velocity adjustment × sampling period. The sampling period is 1 millisecond, corresponding to the CNC system's interpolation period. Position increments are expressed in radians and converted to pulses based on the spindle encoder resolution. The conversion factor is encoder line count / 2π. For an encoder with a resolution of 10,000 lines / rev, the conversion factor is 1591 pulses / radian. Spindle position control is implemented through the CNC system's synchronous control channel. Control commands use the G33.1 format and are dynamically modified using user-defined macros. To prevent command buffer overflow, a preload technique is used to load commands into the cache 20 milliseconds in advance. Phase compensation commands require high real-time performance and are written directly into the spindle drive parameter area using a dedicated communication protocol, with communication latency less than 0.5 milliseconds. The command format includes the execution time (spindle angle trigger), compensation amount (number of pulses), execution duration (angle range), and the final action (restoring standard synchronization or maintaining the new synchronization ratio). The generated spindle instantaneous phase compensation command is transmitted to the execution unit via the CNC system's high-speed bus.

[0103] It is particularly important to calculate the compensation value of the spindle speed and feed axis synchronization ratio based on the predicted pitch deviation in the thread geometry deviation:

[0104] Calculate the flow adjustment coefficient of the coolant injection based on the tooth profile angle deviation in the predicted thread geometric deviation to obtain the cooling flow adjustment instruction;

[0105] Extract tool wear deterioration indications based on real-time abnormal machining status data, and calculate the pressure adjustment coefficient of the coolant injection based on the tooth profile angle offset in the predicted thread geometry deviation to obtain the cooling pressure adjustment instruction;

[0106] The compensation or reduction value of the micro feed rate is calculated based on the tooth profile angle deviation and pitch deviation in the predicted thread geometry deviation to obtain the feed fine adjustment amount;

[0107] A cooling feed adjustment command is generated based on the cooling flow rate adjustment command, the cooling pressure adjustment command, and the feed fine adjustment amount.

[0108] In an embodiment of the present invention, the coolant injection flow adjustment coefficient is calculated based on the tooth profile angle offset data in the predicted thread geometry deviation. The standard coolant flow rate is set to 18 liters / minute, and is monitored in real time by a flow meter. The cooling system uses a high-precision variable frequency pump, model Grundfos CM10-2, with a flow adjustment range of 5-30 liters / minute and an adjustment accuracy of 0.2 liters / minute. The tooth profile angle offset data is converted into a flow adjustment coefficient through linear mapping, and the conversion relationship is: every 0.1 degree angle offset corresponds to a 1.2 liter / minute flow adjustment. For the predicted angle offset value of 0.35 degrees, the calculated flow increase value is 4.2 liters / minute. When the angle offset exceeds 0.5 degrees, the flow increase value is capped at 6 liters / minute to prevent excessive coolant impact on the workpiece. A temperature correction factor is introduced. When the cutting zone temperature (measured by an infrared thermometer) increases by 10°C, the flow rate increases by an additional 0.8 liters / minute. The system sends flow adjustment commands to the variable frequency pump controller via the Modbus protocol. The command format is a 16-bit unsigned integer with a unit of 0.1 liters / minute and a transmission cycle of 100 milliseconds. During command execution, a gradient adjustment strategy is used to smoothly transition to the target flow value within 500 milliseconds to avoid hydraulic shock. The tool wear deterioration index (DWI) is calculated by integrating three indicators: the sudden change rate of the cutting force, the increase in the high-frequency component of the force spectrum, and the periodic change of force fluctuations. The DWI value ranges from 0 to 100, and a value exceeding 65 indicates accelerated wear deterioration. The coolant pressure adjustment coefficient is calculated based on the tooth profile angle deviation in the predicted thread geometry deviation. The standard coolant pressure is 2.0 MPa, and the pressure regulation device is a proportional pressure reducing valve (FESTO MPPES-3-1 / 8-6-010) with an adjustment range of 0.5-6 MPa and a response time of 25 milliseconds. Pressure adjustment is calculated using a piecewise function: when the DWI is below 65, each 0.1-degree angular deviation corresponds to a 0.15 MPa pressure increase; when the DWI is above 65, each 0.1-degree angular deviation corresponds to a 0.25 MPa pressure increase. For a DWI of 78 and a 0.35-degree angular deviation, the calculated pressure increase is 0.875 MPa. When the calculated pressure exceeds the system's maximum safety pressure of 5.5 MPa, a safety limit is activated, limiting it to 5.5 MPa. Pressure adjustment commands are transmitted to the hydraulic control unit via the industrial fieldbus. These commands include the target pressure value, execution time, and pressure hold duration. Control is performed using the Siemens CNC system's F command, with a minimum resolution of 0.001 mm / min. Feed fine-tuning calculations are divided into two parts: pitch deviation-based adjustment and thread angle deviation-based adjustment. Pitch deviation is converted into a feed rate adjustment factor: every 0.01 mm pitch deviation corresponds to a 0.5% feed rate adjustment. For a pitch deviation value of -0.015 mm (the actual pitch is less than the standard value), it is calculated that the feed rate needs to be reduced by 0.75%, that is, by 0.45 mm / min.The tooth profile angle deviation is converted into an additional feed rate adjustment using the following equation: every 0.2-degree angle deviation corresponds to a 0.3% feed rate adjustment. For an angle deviation of 0.35 degrees, the calculated feed rate reduction is 0.525%, or 0.315 mm / min. The combined total feed fine-tuning is -0.765 mm / min. The system sets the upper limit for feed adjustment to ±3%. If this limit is exceeded, an alarm is issued and an exception log is recorded. Feed fine-tuning commands are updated in real time via G01 F commands. Tool life is assessed before execution. When the remaining tool life falls below 20%, the maximum allowable feed rate is automatically reduced by 5%. A parametric macro program, numbered O9050, is used to implement multi-parameter coordinated control. The macro program is stored in the CNC system's user program area. The command generation process begins with a parameter compatibility check to ensure that all parameters are adjusted in the same direction. When the coolant flow rate is increased and the feed rate is decreased, the execution interval between the two is increased. This interval is set to 300 milliseconds to prevent workpiece thermal deformation caused by overcooling. The instruction priority is set as: cooling pressure adjustment (highest), cooling flow adjustment (medium), feed rate adjustment (lowest). After the high priority instruction is executed, the system waits for 25 milliseconds before executing the next priority instruction. The instruction format uses a combination of G code and M code: cooling pressure adjustment uses M138 P value instruction; cooling flow adjustment uses M139 Q value instruction; feed rate adjustment uses G01 F value instruction. The three instructions are combined into an execution block, and the execution time point is specified by the R parameter. The system sets the instruction execution condition judgment. When the tool temperature (measured by the embedded thermocouple) exceeds 320°C, the cooling pressure and flow are automatically adjusted to the maximum allowable values, and the feed rate is reduced to 85% of the standard value. The instruction execution results are fed back to the monitoring system in real time, and the current execution status is displayed on the touch screen.

[0109] The present invention uses a high-frequency three-axis force sensor to collect cutting force data in real time and construct a three-dimensional vector trajectory, and combines the top cutting force density analysis and the tooth side cutting force component extraction to evaluate the thread tooth profile characteristics; identifies abnormal situations in the cutting process through periodic peak envelope processing and dynamic mean trajectory analysis; uses a high-precision three-dimensional coordinate measuring machine to scan the thread surface and analyze the pitch expansion and contraction, effective diameter offset, tooth profile angle distortion and top fillet curvature change; judges the tool wear mode and predicts the thread geometry deviation based on the abnormal situation and precision deviation value; and calculates the spindle phase compensation, cutting depth adjustment and cooling parameter control instructions in a targeted manner to realize the coordinated adaptive control of multiple parameters in the thread processing process, so as to ensure the stable and controllable quality of the fasteners.

[0110] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0111] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling process parameters in fastener production, characterized in that: The following steps are involved: Step S1: Real-time acquisition of the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining, dynamic trajectory tracking, and generation of three-dimensional vector trajectory data of the cutting force; thread profile feature analysis based on the three-dimensional vector trajectory data of the cutting force to generate thread profile cutting force feature data; correlation of thread turn fluctuation amplitude based on the thread profile cutting force feature data to generate dynamic fluctuation feature data of the cutting force; Step S2: performing a cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data; Step S3: Scan the contour of the local area of ​​the machined thread surface of the fastener, calculate the local deviation value of the thread, and generate the local precision deviation value of the thread; judge the tool wear pattern based on the real-time abnormal machining status data and the local precision deviation value of the thread, and predict the thread circle machining deviation to generate the predicted thread geometry deviation; Step S4: Adjust multiple processing control parameter values ​​according to the predicted thread geometry deviation to generate a composite thread processing adjustment instruction; issue a real-time execution control instruction to the machine tool based on the composite thread processing adjustment instruction, and perform real-time processing feedback monitoring to achieve adaptive control of fastener production.

2. The method for controlling process parameters in fastener production according to claim 1, characterized in that: In step S1, the three-dimensional cutting force of the tool-workpiece contact area during the fastener thread machining process is collected in real time, and the trajectory is dynamically tracked, including: The high-frequency three-dimensional force sensor installed on the machine tool turret is used to synchronously collect the radial cutting force, axial cutting force and tangential cutting force in the tool-workpiece contact area during the fastener thread machining process, and obtain instantaneous three-dimensional cutting force time series data; The radial component, axial component and tangential component in the instantaneous three-dimensional cutting force time series data are calculated at each digital sampling point to obtain the instantaneous series data of the cutting force amplitude energy. Perform square root operation on instantaneous series data of cutting force amplitude energy to obtain instantaneous synthetic cutting force data; The three-dimensional vector trajectory of the cutting force is dynamically tracked according to the instantaneous synthetic cutting force data to generate the three-dimensional vector trajectory data of the cutting force.

3. The method for controlling process parameters in fastener production according to claim 1, characterized in that: The thread profile evaluation based on the cutting force three-dimensional vector trajectory data in step S1 includes: Based on the three-dimensional vector trajectory data of cutting force, the fastener thread spatial area is mapped to generate the thread cutting force spatial distribution data; According to the spatial distribution data of thread cutting force, the cutting force density of the thread top fillet radius area in the fastener is analyzed to obtain the top cutting force density data; Extract the cutting force component of the thread tooth flank inclination angle area in the fastener based on the thread cutting force spatial distribution data to obtain the tooth flank cutting force component data; The thread profile geometry coupling analysis is performed based on the tooth top cutting force density data and the tooth side cutting force component data to obtain the thread forming mechanical distribution data; Obtain fastener effective diameter size constraint data; The radial force concentration analysis of the thread forming mechanical distribution data is carried out through the effective diameter size constraint data of the fastener, and then the thread tooth profile angle forming mechanical evaluation is carried out to generate the thread tooth profile cutting force characteristic data.

4. The method for controlling process parameters in fastener production according to claim 1, characterized in that: In step S1, correlating the thread circle fluctuation amplitude according to the thread profile cutting force characteristic data includes: The thread profile cutting force characteristic data is periodically segmented according to the preset fastener thread pitch standard spacing to obtain the single pitch cutting force cycle data; According to the single pitch cutting force cycle data, the average value of each direction force in each thread pitch processing cycle is calculated to obtain the cycle average cutting force vector data; According to the single pitch cutting force cycle data, the peak values ​​of the forces in each direction in each thread pitch processing cycle are extracted to obtain the cycle peak cutting force vector data; The cycle peak cutting force vector data is processed into a dynamic peak envelope, and the cycle average cutting force vector data is processed into a dynamic mean trajectory. Then, the ratio of the instantaneous cutting force peak to the mean in each direction is calculated to generate the dynamic fluctuation coefficient of the force ratio. Evaluate the cutting process stability index based on the dynamic fluctuation coefficient of force ratio; Based on the cutting process stability index, the thread profile cutting force characteristic data is correlated with the thread circle fluctuation amplitude to generate the cutting force dynamic fluctuation characteristic data.

5. The method for controlling process parameters in fastener production according to claim 4, characterized in that: Step S2 includes the following steps: Step S21: performing cutting process vibration suppression analysis based on the cutting force dynamic fluctuation characteristic data, and performing fluctuation-tool wear sensitivity matching to generate a tool wear indicator factor; Step S22: obtaining reference average cutting force vector data and reference cutting force peak-to-average ratio data based on the tool wear indicator factor; Step S23: subtracting the corresponding components of the reference average cutting force vector data from the period average cutting force vector data to obtain average force deviation vector data; Step S24: subtracting corresponding items from the reference cutting force peak-to-average ratio data using the force ratio dynamic fluctuation coefficient to obtain a peak-to-average ratio deviation sequence; Step S25: Identify the real-time abnormal situation of the current machining process based on the average force deviation vector data and the peak-to-average ratio deviation sequence to obtain real-time abnormal machining situation data.

6. The method for controlling process parameters in fastener production according to claim 1, characterized in that: Scanning the local area profile of the machined thread surface of the fastener and calculating the local thread deviation value in step S3 includes: Scan the contour of the local area of ​​the processed thread surface of the fastener to generate the local 3D topography point cloud data of the thread; The fastener's spatial position is synchronized based on the local 3D thread topography point cloud data. Local fitting is then performed on the instantaneous distance change of continuous tooth tops or roots along the axial direction, the radial position of the middle point of the tooth profile, the instantaneous slope of the tooth profile side profile, and the tooth top arc shape to obtain the instantaneous thread geometric characteristic parameter data. Among them, the instantaneous thread geometric characteristic parameter data includes the instantaneous expansion and contraction of the pitch, the instantaneous radial offset of the thread effective diameter, the instantaneous distortion of the thread profile angle, and the instantaneous curvature change of the thread top fillet radius. Based on the instantaneous geometric characteristic parameter data of the thread and the preset target geometric characteristic value of the thread, a real-time deviation calculation is performed to obtain the instantaneous geometric deviation data of the thread; When the instantaneous expansion and contraction of the thread pitch in the instantaneous geometric deviation data shows a negative cumulative trend for three consecutive pitches, it is determined to be in the state of cumulative pitch contraction; When the instantaneous distortion of the thread profile angle in the thread instantaneous geometric deviation data continuously exceeds the preset angle deviation value threshold of 0.5 degrees, it is determined to be a thread profile angle distortion state; When the instantaneous radial deviation of the effective diameter of the thread in the instantaneous geometric deviation data exceeds 0.005 mm continuously, the effective diameter deviation state is obtained; When the instantaneous curvature change of the thread top fillet radius in the thread instantaneous geometric deviation data continuously exceeds the preset relative reference curvature change by 5%, the thread top fillet curvature deviation state is obtained; Dynamically assign weights to the instantaneous geometric deviation data of the thread to obtain the thread local deviation weight data; The thread local deviation weight data is used to perform weighted aggregation calculation on the cumulative state of pitch shrinkage, tooth profile angle distortion, effective diameter offset and top fillet curvature deviation to generate the thread local accuracy deviation value.

7. The method for controlling process parameters in fastener production according to claim 1, characterized in that: In step S3, the tool wear pattern is judged based on the real-time abnormal machining situation data and the thread local precision deviation value, and the thread circle machining deviation is predicted, including: The tool wear pattern is judged based on the local thread accuracy deviation value and real-time abnormal processing status data to generate tool wear pattern recognition data. The tool wear pattern judgment includes the tool tip / side edge grinding circular wear pattern and the main cutting edge local chipping wear pattern; The wear degree of the tool wear pattern recognition data is quantified by the local thread accuracy deviation value, and the wear value of the tool tip radius and the wear degree of the side edge are estimated respectively. Obtain the maximum tool tip radius wear and maximum side edge wear allowed by tool design; According to the maximum tool tip radius wear and the maximum side edge wear, the tool tip radius wear estimation and side edge wear estimation are calculated respectively. The larger value of the two is taken as the tool comprehensive wear percentage. Based on the tool comprehensive wear percentage, the tool wear pattern recognition data is used to predict the thread circle diameter deviation, pitch deviation and tooth profile angle deviation to obtain the predicted thread geometry deviation.

8. The method for controlling process parameters in fastener production according to claim 1, characterized in that: In step S4, adjusting multiple processing control parameter values ​​according to the predicted thread geometry deviation includes: The compensation value of the spindle speed and feed axis synchronization ratio is calculated based on the pitch deviation in the predicted thread geometric deviation, and the instantaneous phase compensation instruction of the spindle is obtained; Calculate the adjustment value of the radial cutting depth according to the thread circle diameter deviation in the predicted thread geometric deviation to obtain the cutting depth adjustment instruction; Based on the real-time abnormal machining situation data, the adjustment value of the cooling parameter is calculated based on the tooth profile angle offset in the predicted thread geometry deviation, and the cooling feed adjustment instruction is obtained; The spindle instantaneous phase compensation instruction, cutting depth adjustment instruction and cooling feed adjustment instruction are combined into processing control parameters to obtain the compound thread processing adjustment instruction.

9. The method for controlling process parameters in fastener production according to claim 8, characterized in that: The compensation values ​​for calculating the synchronization ratio of the spindle speed and the feed axis based on the pitch deviation in the predicted thread geometry deviation include: According to the pitch deviation in the predicted thread geometry deviation, the direction and value of the pitch cumulative error are identified, and the single-turn pitch compensation amount is evaluated to obtain the single-turn pitch compensation amount data; Get the current spindle speed and thread lead data; Based on the single-turn pitch compensation data, the instantaneous angular velocity adjustment amplitude that the spindle and the feed axis should perform in each thread spiral turn is calculated through the current spindle speed and thread lead data to obtain the spindle angular velocity adjustment amplitude data; Taking every 30 degrees as an adjustment point, the spindle angular velocity adjustment amplitude data is decomposed into a sequence of instantaneous angular velocity adjustment amounts on the thread cutting path to generate the spindle instantaneous angular velocity adjustment data; The instantaneous angular velocity of the main shaft is adjusted according to the data of the angular velocity modulation instruction, and the instantaneous phase compensation instruction of the main shaft is generated.

10. A process parameter control system for fastener production, characterized in that: The method for controlling process parameters in fastener production according to claim 1 is used to execute the method for controlling process parameters in fastener production according to claim 1, and the control system for controlling process parameters in fastener production comprises: The fastener cutting force analysis module is used to collect the three-dimensional cutting force in the tool-workpiece contact area during fastener thread machining in real time, and dynamically track the trajectory to generate three-dimensional cutting force vector trajectory data; based on the three-dimensional cutting force vector trajectory data, it analyzes the thread profile characteristics to generate thread profile cutting force characteristic data; based on the thread profile cutting force characteristic data, it correlates the thread circle fluctuation amplitude to generate cutting force dynamic fluctuation characteristic data; The machining anomaly monitoring module is used to perform vibration suppression analysis during the cutting process based on the dynamic fluctuation characteristic data of the cutting force, so as to identify the real-time abnormal situation of the current machining process and obtain real-time abnormal machining situation data; The accuracy deviation prediction module is used to scan the local area profile of the machined thread surface of the fastener, calculate the local thread deviation value, and generate the local thread accuracy deviation value; based on the real-time abnormal processing status data and the local thread accuracy deviation value, it determines the tool wear pattern and predicts the thread circle processing deviation to generate the predicted thread geometry deviation; The adaptive production control module is used to adjust the values ​​of multiple processing control parameters based on the predicted thread geometry deviation and generate compound thread processing adjustment instructions; based on the compound thread processing adjustment instructions, the machine tool executes control instructions in real time and performs real-time processing feedback monitoring to achieve adaptive control of fastener production.

Images