Bolt tightening torque control method and system based on yield point method

Abstract

The invention discloses a bolt tightening torque control method based on a yield point method, and relates to the technical field of bolt tightening. Calculating the slope differential quotient of the torque-rotation angle curve; a machine learning model is utilized to adapt to a specified value range, and ultrasonic pre-tightening force data are fused to carry out cross validation to obtain an adjusted specified value range; and when the slope differential quotient is reduced to an adjusted specified value range, judging that the bolt reaches a yield point, and sending a tightening stopping signal. According to the method, the slope differential quotient is calculated in real time, it is judged that the bolt reaches the yield point when the slope differential quotient is lowered to be within the regulated value range after adjustment, stable control over the pre-tightening force is achieved, inapplicability of a traditional torque / corner method to a small-corner short bolt is avoided, the risk of plastic deformation of a bolt thread and a rod part is reduced, reusability and durability are improved, and the method is suitable for popularization and application. The bolt is not influenced by a friction coefficient and a corner starting point, and material fatigue and loosening risks are reduced on occasions where the bolt is poor in plasticity or repeatedly used.

Description

Technical Field

[0001] This invention relates to the field of bolt assembly technology, and in particular to a bolt tightening torque control method based on the yield point method. Background Technology

[0002] Bolt tightening is a common fastening process in mechanical assembly and engineering structure installation. Its core lies in ensuring the reliability and tightness of the connection by applying appropriate preload, preventing gaps or relative slippage under working loads. Currently, the main method for controlling bolt tightening torque is the torque / angle method, which is widely used in industrial automation equipment. This method first tightens the bolt to a preset torque value, and then continues to rotate the nut to a specified angle to indirectly control the preload. Specifically, refer to... Figure 2 , Figure 3 and Figure 4 Within the elastic zone, the relationship between the preload F and the rotation angle θ can be expressed as F = CaPθ / 360°, where Ca is the system stiffness and P is the thread pitch. In the plastic zone, the threaded fastener is tightened above its yield point, undergoing permanent deformation. At this point, the preload is related to the bolt's yield strength. This method has the advantage of relatively stable tightening quality, and the thread friction coefficient has a relatively small impact on the preload during the rotation angle control stage, because within the elastic deformation zone, the preload mainly depends on the proportional relationship between bolt elongation and rotation angle.

[0003] While the torque / angle method improves assembly efficiency to some extent, it has several drawbacks: First, it is unsuitable for short bolts with small rotation angles because the preload is often large, leading to plastic deformation of the bolt threads and shank, affecting the bolt's reusability and durability. Second, in situations where bolts have poor plasticity or are used repeatedly, this method can easily lead to material fatigue or loosening risks. Furthermore, although changes in the coefficient of friction affect the starting point of the rotation angle, the overall accuracy is limited by the accuracy of the initial torque, failing to fully utilize the bolt's load-bearing potential. These shortcomings are particularly prominent in high-precision assembly scenarios, such as tension clamp bolt tightening, potentially leading to uneven preload distribution and consequently affecting the connection's resistance to loosening and fatigue. Summary of the Invention

[0004] The purpose of this invention is to provide a bolt tightening torque control method based on the yield point method, addressing the shortcomings of existing torque / rotation methods in terms of accuracy, applicability, and robustness. This method achieves stable control of high-precision preload, suitable for complex working conditions such as bolt tightening robots in assembled steel structures. Furthermore, this invention provides a dynamic adjustment mechanism within a specified value range by integrating machine learning models for adaptive optimization and fusing ultrasonic preload data for cross-validation, further enhancing its anti-interference capability under complex working conditions.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide a bolt tightening torque control method based on the yield point method, comprising: real-time acquisition of torque and angle values ​​during the tightening process; calculation of the slope derivative of the torque-angle curve; adaptive adaptation of a specified value range using a machine learning model, and cross-validation by integrating ultrasonic preload data to obtain an adjusted specified value range; determination of whether the slope derivative has decreased from its peak value to the adjusted specified value range; and determination that the bolt has reached its yield point and issuance of a stop tightening signal when the slope derivative value decreases to the adjusted specified value range.

[0006] As a preferred embodiment of the bolt tightening torque control method based on the yield point method described in this invention, the expression for the slope derivative is: ; In the formula, This represents the change in tightening torque. This represents the change in the torsion angle.

[0007] As a preferred embodiment of the bolt tightening torque control method based on the yield point method described in this invention, the step of using a machine learning model to adaptively determine the specified value range and fusing ultrasonic preload data for cross-validation includes: collecting historical datasets, including torque values, rotation angle values, slope derivatives, actual yield point locations, and related features; preprocessing and labeling the targets; initializing the machine learning model and performing initial training; real-time acquisition of torque values, rotation angle values, and ultrasonic preload data; calculation of slope derivatives; monitoring of peak values ​​and extraction of dynamic features; inputting real-time features into the machine learning model for adaptive calculation of the specified value range; updating the model weights using an online learning mechanism; and applying a smoothing filter to process the specified value range; fusing ultrasonic preload data for cross-validation; adjusting the specified value range by comparing the consistency of the determinations and using a weighted average method; applying the adjusted specified value range to the judgment module; when the slope derivative enters the adjusted specified value range and remains there for more than a time judgment window, determining that the yield point has been reached and issuing a stop signal; simultaneously recording data for model training; and reverting to the default specified value range in abnormal situations. The default value range is set to 1 / 2. Maxed to 1 / 3 max, where `max` represents the maximum slope of the elastic zone.

[0008] As a preferred embodiment of the bolt tightening torque control method based on the yield point method described in this invention, the method is applied to the bolt tightening process and includes four stages: Thread contact and initial alignment stage: The drive motor begins to tighten, the friction is small, and the torque and angle are monitored in real time, but the slope quotient is close to zero and is not used as a control basis; During the thread engagement and friction overcoming stage: a small torque is applied to overcome friction, the slope derivative gradually increases, and the area is in the elastic zone; Flange contact and friction peak stage: The frictional force reaches its maximum, the slope derivative reaches its peak and then begins to decrease, and when the slope derivative first decreases to the adjusted specified value range, the bolt is judged to have reached the yield point; Tightening completion and yield point control stage: The torque remains constant, the rotation angle continues to increase, and the slope derivative is maintained within the adjusted specified value range for more than the time judgment window. When this period exceeds the time judgment window, the system issues a control signal to end the tightening work and complete the assembly.

[0009] As a preferred embodiment of the bolt tightening torque control method based on the yield point method of the present invention, the time determination window is set to a range of 50-100 milliseconds.

[0010] As a preferred embodiment of the bolt tightening torque control method based on the yield point method of the present invention, the method further includes analyzing the relationship between the bolt thread preload and the tightening torque. The tightening torque expression is: T = T1 + T2, where T is the tightening torque, T1 is the thread torque between the screw pairs, and T2 is the friction torque on the nut support surface.

[0011] As a preferred embodiment of the bolt tightening torque control method based on the yield point method described in this invention, the thread preload F of the bolt is linearly related to the tightening torque T until the curve bends downward after exceeding the yield point.

[0012] Secondly, to further address the problems existing in bolt tightening, the present invention provides a bolt tightening torque control system based on the yield point method, comprising: a data acquisition module for real-time acquisition of torque and angle values ​​during the tightening process; a calculation module for calculating the slope derivative of the torque-angle curve; and an adaptive specified value module for collecting historical datasets for machine learning model initialization and training, real-time acquisition of torque, angle, and ultrasonic preload data to extract dynamic features, inputting real-time features into the model for adaptive calculation and adjustment of the specified value range, updating model weights using an online learning mechanism, and applying a smoothing filter to process the specified value. The system employs a multi-factor algorithm, which integrates ultrasonic preload data for cross-validation. An adjusted specified value range is generated using a weighted average method and fed back to the judgment module. Simultaneously, data is recorded for subsequent training, and a default specified value range is reverted in case of anomalies. The judgment module determines whether the slope derivative has decreased from its peak value to within the adjusted specified value range. The control module determines that the bolt has reached its yield point and issues a stop-tightening signal when the slope derivative value decreases to within the adjusted specified value range. The simulation analysis module simulates the bolt tightening process, monitors the dynamic relationship between motor speed and output torque, and divides the process into four tightening stages.

[0013] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, implements any step of the bolt tightening torque control method based on the yield point method as described in the first aspect of the present invention.

[0014] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the bolt tightening torque control method based on the yield point method as described in the first aspect of the present invention.

[0015] The beneficial effects of this invention are as follows: This invention collects torque and angle values ​​in real time, calculates the slope derivative of the torque-angle curve, and determines that the bolt has reached the yield point when the slope derivative drops from the peak value to a set specified value range. This achieves high-precision stable control of preload, avoids the inapplicability of the traditional torque / angle method to short bolts with small angles, reduces the risk of plastic deformation of the bolt thread and shank, and improves the reusability and durability of the bolt. This invention is not affected by the coefficient of friction and the starting point of the rotation angle, and can overcome the defects of the torque control method and the elastic zone rotation angle method. In situations where the bolt has poor plasticity or is used repeatedly, it can effectively reduce the risk of material fatigue or loosening, ensure uniform distribution of preload, and improve the anti-loosening and anti-fatigue performance of the connecting parts. This invention improves the robustness and precision adaptability of the bolt tightening process by using a dynamic adjustment mechanism within a specified value range, employing a machine learning model to adaptively optimize the specified value range, and integrating ultrasonic preload data for cross-validation. It can respond in real time to complex working conditions such as fluctuations in friction coefficient, material variations, or environmental interference, avoiding misjudgments or precision decay caused by a fixed specified value range. This significantly improves assembly efficiency, reduces the need for human intervention, and ensures that the preload is controlled at the maximum strength potential near the yield point, thus extending the overall service life of the connecting parts. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a flowchart illustrating the implementation of the method in this invention.

[0017] Figure 2 This is the curve showing the relationship between preload and tightening torque in this invention.

[0018] Figure 3 This is a diagram showing the relationship between the rotation angle of the elastic region and the preload in this invention.

[0019] Figure 4 This is a diagram showing the relationship between the rotation angle of the plastic region and the preload in this invention.

[0020] Figure 5 This is a schematic diagram of the yield point method in this invention.

[0021] Figure 6 This is a dynamic relationship diagram of motor speed and output torque during the bolt tightening process in this invention. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Example 1 The following is the first embodiment of the present invention, which provides a bolt tightening torque control method based on the yield point method, including: The first part analyzes the relationship between the thread preload and tightening torque of the bolt, specifically: Bolt threaded connections must be tightened during assembly to pre-apply force before bearing working loads; this pre-force is called preload F. The purpose of pre-tightening is to enhance the reliability and tightness of the connection, preventing gaps or relative slippage after loading. When tightening a nut, the main force overcomes the thread torque T1 between the screw threads and the frictional torque T2 on the nut's supporting surface; therefore, the tightening torque T = T1 + T2. Taking an M12 bolt on a tension clamp as the research object, and temporarily disregarding bolt deformation and corrosion, the tightening torque T follows a predictable pattern. Experiments have shown that the preload F of an M12 bolt thread and the tightening torque T theoretically have a linear relationship, such as... Figure 2 As shown, when the preload exceeds the yield point, the increment of the preload decreases. When the tightening torque exceeds 80 N·m, the relationship curve shows a downward curve shape.

[0026] Part Two: Methods for controlling the tightening torque of bolt threads, specifically: For the tightening of tension clamp bolts, the essence of ensuring tight bolt threads is to control the preload within an appropriate range. Research and analysis show that parameters such as torque and rotation angle are related to the preload; controlling these parameters can indirectly control the preload. In this embodiment, the thread tightening control method is the yield point method.

[0027] The yield point method involves monitoring the slope of the tightening torque versus angle curve to determine the thread level of the threaded component. The principle of the yield point method is described below. Figure 5As shown, the F curve represents the preload, and the T curve represents the tightening torque. During thread tightening, the computer can calculate the slope of the torque-angle curve at any time. When the slope is horizontal, both the torque and preload are in the elastic zone of tightening; when the slope drops significantly, it indicates that the material has entered the plastic zone. From the elastic zone to the plastic zone, the linear relationship between torque and angle changes, and the slope also changes. When the slope reaches a certain range, it is considered to have reached the yield point. In the assembly process using an intelligent control tightening device, not only can the assembly torque of the threaded pair be tested, but the angle can also be continuously tested simultaneously with the torque test, thereby calculating the derivative of the slope between torque and angle. The formula for calculating the derivative of the slope is as follows: ; In the formula, This represents the change in tightening torque. This refers to the change in torsion angle. When the bolt material reaches its yield point, i.e., when the tightening torque no longer increases or increases slowly, but the torsion angle increases rapidly, at this point... When the value approaches the specified range, typically (1 / 2 to 1 / 3) of Δmax, a control signal is issued to end the tightening process, completing one work cycle. Here, Δmax represents the maximum slope of the elastic zone. Therefore, the yield point method utilizes the concept of torque-angle increment ratio to tighten threaded parts to the bolt's yield point. This method is unaffected by the friction coefficient and angle starting point of the torque / angle control method, thus overcoming the shortcomings of the torque control method and the elastic zone angle method. It improves assembly accuracy and, by tightening the bolt to its yield point, maximizes the potential strength of the threaded parts, enhancing resistance to loosening and fatigue.

[0028] The specified value range can be dynamically adjusted, and the specific implementation steps are as follows: Data preparation and model initialization: Historical datasets were collected from previous bolt tightening experiments or production data, including torque value T, rotation angle value θ, and calculated slope derivative. First, confirm the actual yield point location and related characteristics, such as bolt type, material yield strength, friction coefficient, and temperature, through laboratory testing or simulation. Preprocess the data, including cleaning outliers, normalizing features, and labeling the target, i.e., the optimal specified value range, for example, an adaptive value of 1 / 2 to 1 / 3 of the historical Δmax. Initialize the machine learning model, choosing a Support Vector Regression (SVR) or Neural Network (such as MLP) model. The model input consists of real-time relevant features and historical data, and the output is the lower and upper limits of the adjusted specified value range. Initial training uses supervised learning, employing mean squared error as the loss function to minimize the deviation between the predicted specified value and the actual yield point.

[0029] Real-time data acquisition and feature extraction: During the tightening process, torque and angle values ​​are acquired in real time using torque sensors and angle encoders. Simultaneously, ultrasonic sensors are deployed to measure the preload force F, serving as auxiliary data for cross-validation. The slope of the torque-angle curve is continuously calculated based on the slope derivative formula. The maximum slope Δmax in the elastic zone is monitored. Dynamic features are extracted from the real-time data, such as the current frictional torque T2, the linear relationship between the preload force and the preload (the bending point of the curve between F and T), changes in motor speed (refer to the simulation analysis module), and external variables (e.g., the effect of ambient humidity on friction).

[0030] Adaptive Adjustment of Machine Learning Models: Real-time features are input into the trained machine learning model. The model adaptively calculates new lower and upper limits for the specified value range based on the input. For example, if an increase in the friction coefficient is inferred from detected ultrasonic data, the lower limit of the specified value is adjusted from 1 / 2Δmax to a higher value (e.g., 0.55Δmax) to avoid premature judgment. Online learning mechanisms, such as incremental training, are used to update the model weights after each batch of tightening data. If historical data indicates that certain bolt types (e.g., M12 bolts) experience a faster slope decrease in the plastic zone, the range is dynamically reduced, for example, from 1 / 2~1 / 3Δmax to 1 / 2.5~1 / 3.5Δmax. Moving averages or Kalman filters are applied to smoothly adjust the specified value range, avoiding drastic changes caused by noise.

[0031] Cross-validation using ultrasonic preload data: The axial preload F of the bolt is measured in real-time using an ultrasonic sensor, and its relationship with torque is calculated (referencing T = T1 + T2, F and T are linear until the yield point bends downwards). The preload measured by ultrasonic waves is compared with the judgment based on the slope derivative (Δ decreases to the adjusted specified range): If the ultrasonic reading shows F approaching the yield strength, for example, the curve begins to bend downwards, and this is consistent with the Δ judgment, then the adjusted specified range is confirmed to be valid; if inconsistent, for example, the ultrasonic reading shows F has not reached the yield point, but Δ has decreased, then feedback is sent to the model, triggering a readjustment of the specified value, for example, increasing the range to delay the stop signal. The two data sources are fused using a weighted average or Bayesian method, for example, the new specified range = α * Δ. threshold +(1-α)* F threshold Where α is the weight, initially set to 0.7 and adjusted based on the verification error, Δ threshold F represents the range of values ​​specified for determining the derivative based on the slope. threshold This indicates the specified range of values ​​based on ultrasonic preload data.

[0032] Judgment and Control Execution: The adjusted specified value range is applied to the judgment module. When the slope derivative decreases from the peak and enters the adjusted specified value range, and remains there for more than the time judgment window (50-100ms), the yield point is determined to have been reached. The control module issues a stop tightening signal, ending the process. Simultaneously, this data is recorded for subsequent model training. If the adjustment fails (e.g., low model confidence), it reverts to the default specified value range (1 / 2~1 / 3Δmax) and logs a warning.

[0033] Example 2 The following is a second embodiment of the present invention. This embodiment differs from the first embodiment in that it also provides a simulation analysis of a bolt thread tightening torque control method. It utilizes a bolt tightening torque control method based on the yield point method, employing constant power control for the bolt tightening motor. For a fully loose bolt, the entire tightening process is divided into four stages, and the tightening torque T and torsion angle θ are monitored in real time at each stage. The slope derivative of the torque-torsion angle curve is calculated. This is to determine whether the bolt has reached its yield point. The specific implementation process is as follows: The threaded contact and initial alignment stage lasts 0-700ms: When the drive motor begins tightening, the bolt and nut threads make contact. During this stage, the friction between them is low, and the main purpose is to move the nut to a specified angle to ensure alignment between the nut and bolt and avoid errors. The motor output torque is approximately zero, and the motor speed is close to the no-load speed. The tightening torque T and torsion angle θ are monitored in real time, but due to the extremely small torque change, the slope is minimal. Values ​​close to zero are not used as a basis for control decisions.

[0034] The thread engagement and friction-overcoming phase lasts 700-2000 ms: The bolt and nut threads come into contact and partially engage. At this point, a small torque is applied to overcome the friction caused by their contact. The motor output torque gradually increases, while the motor speed relatively decreases. Real-time calculation of the slope derivative is then performed. ,at this time The value gradually increases, indicating that the system is in the elastic zone, and the torque and rotation angle have a linear relationship. This stage mainly ensures smooth thread engagement and does not trigger the yield point judgment.

[0035] During the flange contact and friction peak phase, lasting 2000-3300 ms, the nut moves smoothly until it contacts the flange, at which point the maximum and most stable frictional force appears. The motor output torque gradually increases to its maximum value, while the motor speed gradually decreases to near zero. During this phase, it is necessary to monitor the angular displacement of the nut to provide feedback on the distance the nut has traveled along the bolt axis, which helps in estimating the effective bolt length. Real-time calculation of the slope derivative is also crucial. ,when The value reached its peak. The material begins to decrease after reaching its maximum value, indicating that it is entering the plastic zone. When... The value first drops to the specified range (usually 1 / 2). Maxed to 1 / 3 When the bolt reaches its maximum yield point (max), the system considers the bolt to have reached its yield point and prepares to issue a control signal.

[0036] The tightening completion and yield point control stage takes 3300-4200ms: the bolt is tightened, the nut has reached the flange, and the final part of the tightening process begins. During the tightening and rotating of the nut, the desired clamping force is generated between the flange and the nut. The motor output torque remains at a constant maximum value, and the motor speed is 0, creating a temporary stall state. The slope derivative is calculated in real time. Furthermore, the system sets a short time window for judgment (e.g., 50-100 milliseconds), and since the torque no longer increases while the steering angle continues to increase, The value further decreased, when When the preload remains within the specified range for an extended period, the system issues a control signal to terminate the tightening process and complete the assembly. This method ensures that the preload is controlled near the yield point, fully utilizing the bolt's strength potential and improving its resistance to loosening and fatigue.

[0037] Furthermore, experiments were conducted and analyzed with M12 bolts completely loose and unloaded. A DC24V, 150W Maxon EC-max series motor was selected, and the entire bolt tightening process was simulated and analyzed in CSharp software. The dynamic relationship between motor speed and output torque was referenced... Figure 6 .

[0038] As shown in the dynamic relationship graph of motor speed and output torque, the left vertical axis represents the current value (mA); the right vertical axis represents the number of pulses recorded by the encoder; and the horizontal axis represents the time constant (ms). The straight line in the graph represents the motor speed, and the wavy line represents the motor stator current. Since the motor output torque is proportional to the square of the stator current, changes in the current value on the left reflect changes in the motor output torque, and changes in the number of pulses recorded by the encoder on the right reflect changes in the motor speed. Figure 6In the simulation, [0, 700] ms represents the first stage, where the friction between the bolt and nut is approximately zero, the motor output torque is approximately zero, and the motor speed is close to the no-load speed. [700, 2000] ms represents the second stage, where the bolt and nut threads engage, the motor output torque gradually increases, and the motor speed relatively decreases. [2000, 3300] ms represents the third stage, where the nut gradually reaches and eventually contacts the flange, resulting in a maximum and stable frictional resistance. The motor output torque gradually increases to a maximum value, and the motor speed gradually decreases to near zero. [3300, 4200] ms represents the fourth stage, where the nut has reached the flange, the motor output torque is at a constant maximum value, and the motor speed is zero, creating a temporary stall state. The bolt can be considered tightened, and the motor then stops, ending the tightening operation. The simulation results are consistent with the theoretical analysis, indicating that the adopted control principle and method can complete the bolt tightening operation of the tension clamp.

[0039] Example 3 The following is a third embodiment of the present invention. This embodiment differs from the first embodiment in that it also provides a bolt tightening torque control system based on the yield point method, comprising: The data acquisition module is used to collect torque and angle values ​​in real time during the tightening process; The calculation module is used to calculate the slope derivative of the torque-steering angle curve; The adaptive specified value module is used to collect historical datasets for machine learning model initialization and training. It collects torque, angle and ultrasonic preload data in real time to extract dynamic features, inputs real-time features into the model to adaptively calculate and adjust the specified value range, updates model weights using an online learning mechanism and applies a smoothing filter to process the specified values, fuses ultrasonic preload data for cross-validation, generates the adjusted specified value range through a weighted average method, feeds the adjusted specified value range back to the judgment module, records data for subsequent training, and reverts to the default specified value range in abnormal situations. The judgment module is used to determine whether the slope derivative has decreased from the peak value to within the adjusted specified value range; The control module is used to determine that the bolt has reached the yield point and issue a stop tightening signal when the slope derivative value drops to within the adjusted specified range; The simulation analysis module is used to simulate the bolt tightening process, monitor the dynamic relationship between motor speed and output torque, and divide it into four tightening stages.

[0040] This embodiment also provides a computer device applicable to a bolt tightening torque control method based on the yield point method, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the bolt tightening torque control method based on the yield point method proposed in the above embodiment.

[0041] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0042] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a bolt tightening torque control method based on the yield point method as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0043] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A bolt tightening torque control method based on the yield point method, characterized in that, include: Real-time acquisition of torque and rotation values ​​during bolt tightening; Calculate the slope derivative of the torque-steering angle curve; The adjusted specified value range was obtained by using a machine learning model to adaptively adapt the specified value range and by fusing ultrasonic preload data for cross-validation. Determine whether the slope derivative has decreased from the peak value to the adjusted specified value range; When the slope derivative value drops to the adjusted specified range, it is determined that the bolt has reached the yield point, and a stop tightening signal is issued.

2. The bolt tightening torque control method based on the yield point method as described in claim 1, characterized in that, The expression for the slope derivative is: ； In the formula, This represents the change in tightening torque. This represents the change in the torsion angle.

3. The bolt tightening torque control method based on the yield point method as described in claim 2, characterized in that, The method of using a machine learning model to adaptively determine the specified value range and fusing ultrasonic preload data for cross-validation includes: collecting historical datasets, including torque values, angle values, slope derivatives, actual yield point locations, and related features; preprocessing and labeling the targets; initializing the machine learning model and performing initial training; real-time acquisition of torque values, angle values, and ultrasonic preload data; calculating the slope derivative; monitoring peak values ​​and extracting dynamic features; inputting real-time features into the machine learning model to adaptively calculate the specified value range; updating the model weights using an online learning mechanism; and applying a smoothing filter to process the specified value range; fusing ultrasonic preload data for cross-validation; adjusting the specified value range by comparing the consistency of the determinations and using a weighted average method; applying the adjusted specified value range to the judgment module; when the slope derivative enters the adjusted specified value range and remains there for more than a time judgment window, determining that the yield point has been reached and issuing a stop signal; simultaneously recording data for model training; and reverting to the default specified value range in abnormal situations. The default value range is set to 1 / 2. Maxed to 1 / 3 max, where `max` represents the maximum slope of the elastic zone.

4. The bolt tightening torque control method based on the yield point method as described in claim 1, characterized in that, The method, when applied to the bolt tightening process, comprises four stages: Thread contact and initial alignment stage: The drive motor starts and the tightening process begins. The friction is small. The torque and rotation angle are monitored in real time, but the slope quotient is close to zero and is not used as a control basis. During the thread engagement and friction overcoming stage: a small torque is applied to overcome friction, the slope derivative gradually increases, and the area is in the elastic zone; Flange contact and friction peak stage: The frictional force reaches its maximum, the slope derivative reaches its peak and then begins to decrease, and when the slope derivative first decreases to the adjusted specified value range, the bolt is judged to have reached the yield point; Tightening completion and yield point control stage: The torque remains constant, the rotation angle continues to increase, and the slope derivative is maintained within the adjusted specified value range for more than the time judgment window. When this period exceeds the time judgment window, the system issues a control signal to end the tightening work and complete the assembly.

5. The bolt tightening torque control method based on the yield point method as described in claim 4, characterized in that, The time determination window is set to a range of 50-100 milliseconds.

6. The bolt tightening torque control method based on the yield point method as described in claim 1, characterized in that, The method also includes analyzing the relationship between the thread preload and tightening torque of the bolt; The tightening torque expression is: T = T1 + T2, where T is the tightening torque, T1 is the thread torque between the screw pairs, and T2 is the friction torque on the nut support surface.

7. The bolt tightening torque control method based on the yield point method as described in claim 6, characterized in that, The thread preload F of the bolt is linearly related to the tightening torque T, until the curve bends downward after exceeding the yield point.

8. A bolt tightening torque control system based on the yield point method, comprising the bolt tightening torque control method based on the yield point method according to any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to collect torque and angle values ​​in real time during the tightening process; The calculation module is used to calculate the slope derivative of the torque-steering angle curve; The adaptive specified value module is used to collect historical datasets for machine learning model initialization and training. It collects torque, angle and ultrasonic preload data in real time to extract dynamic features, inputs real-time features into the model to adaptively calculate and adjust the specified value range, updates model weights using an online learning mechanism and applies a smoothing filter to process the specified values, fuses ultrasonic preload data for cross-validation, generates the adjusted specified value range through a weighted average method, feeds the adjusted specified value range back to the judgment module, records data for subsequent training, and reverts to the default specified value range in abnormal situations. The judgment module is used to determine whether the slope derivative has decreased from the peak value to within the adjusted specified value range; The control module is used to determine that the bolt has reached the yield point and issue a stop tightening signal when the slope derivative value drops to within the adjusted specified range; The simulation analysis module is used to simulate the bolt tightening process, monitor the dynamic relationship between motor speed and output torque, and divide it into four tightening stages.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the bolt tightening torque control method based on the yield point method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the bolt tightening torque control method based on the yield point method as described in any one of claims 1 to 7.

Images