An Automatic Adjustment Method for the Riveting Stroke of Clinch Riveting without Nails

Abstract

The present invention discloses a method for automatically adjusting the riveting stroke of non - nail riveting, which relates to the technical field of riveting and is used to solve the problem of riveting precision deviation. First, the theoretical stop position is determined through simulation analysis and experimental data, and a process reference is established based on the riveting process and material characteristics. Then, during the production process, the riveting stop position, ambient temperature, power supply voltage, and vibration data of the workpiece are collected in real - time. The environmental interference is eliminated through a multi - factor correction method, the deviation is calculated and compared with the theoretical parameters to form dynamic compensation. To cope with long - term wear and changes in the installation environment, the present invention also calculates the wear correction amount and geological correction amount through regression analysis and neural network methods, combined with information such as wear data, load history, and geological environment, and dynamically adjusts the riveting stroke according to the corrected deviation. Finally, through real - time feedback and closed - loop control, the riveting process is optimized to achieve precise riveting positioning, improving production efficiency and product quality.

Description

Technical Field

[0001] The present invention relates to the field of riveting technology, and more specifically, to a method for automatically adjusting the riveting-in stroke of nail-free riveting. Background Art

[0002] The riveting process has important applications in fields such as aerospace, automotive manufacturing, and shipbuilding. Especially under the requirements of high-precision and high-strength connections, the riveting quality directly affects the performance and safety of products. Traditional riveting technologies generally rely on manual experience and fixed compensation methods. Although they can meet the needs of some standardized production, in the face of complex situations such as different materials, workpiece shapes, and process fluctuations, their accuracy and consistency have great instability, and it is difficult to ensure the continuous stability of riveting accuracy.

[0003] Among them, the long-term wear of riveting equipment, changes in the use environment (such as temperature, humidity, air flow), and installation foundation (such as geological conditions, soil settlement, etc.) have a particularly prominent impact on riveting accuracy. The existing technologies cannot effectively cope with the dynamic changes of these external and internal factors, resulting in frequent deviations in the riveting stroke, affecting the riveting quality, and further affecting the connection strength and reliability of products. Especially under the requirements of high-precision manufacturing, traditional static compensation methods often cannot adjust the deviations in the riveting process in real time, thus reducing production efficiency and product qualification rate.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for automatically adjusting the riveting-in stroke of nail-free riveting to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] In a preferred embodiment, the method includes the following steps:

[0008] Step 1: Preset the initial riveting-in stroke, calculate the optimal riveting-in stroke, and correct the theoretical stop position;

[0009] Step 2: Collect the actual stop position and data of temperature, power supply, and air flow rate, and preliminarily calculate the deviation;

[0010] Step 3: Calculate the initial compensation value, adjust the stroke parameters of the riveting equipment, and verify the effectiveness of the compensation value;

[0011] Step 4: Calculate and correct the stroke deviation caused by the wear of the riveting equipment and environmental changes, update the compensation value, and perform feedback adjustment.

[0012] In a preferred embodiment, in step 1, a preliminary riveting travel range Pinit is preset. The workpiece, riveting head, and support structure models are drawn using CAD software SolidWorks and imported into FEA software. High-precision meshes are divided. Meanwhile, the displacement constraints of the riveting head are defined, and a riveting pressure Ppress is applied. Finally, through multiple rounds of simulation, the optimal riveting travel Pfea is found.

[0013] Select several workpieces and rivet them respectively at the optimal riveting travel Pfea ± 0.1 mm for testing. Use a servo press for riveting, set the same punch speed and riveting pressure, and eliminate abnormal data that deviate from the mean by ±3σ standard deviation. Calculate the experimental travel Pexp according to the formula:

[0014]

[0015] where N is the number of valid experimental data points, and Pmeasured represents the actual stop position of the workpiece.

[0016] Calculate the corrected theoretical stop position Ptheory according to the formula:

[0017] Ptheory = w1 × Pfea + w2 × Pexp

[0018] where w1 represents the weight of the optimal riveting travel Pfea, and w2 represents the weight of the experimental travel Pexp.

[0019] In a preferred embodiment, in step 2, use a laser displacement sensor to collect the actual stop position Pmeasured of the workpiece, use a platinum resistance temperature sensor to collect ambient temperature data, use a power quality monitoring riveting device to collect the current power supply voltage V and frequency f, and use an IEPE vibration sensor to collect the current vibration value V.

[0020] Use a multi-factor correction method to eliminate the influence of temperature, power supply, and vibration on the actual stop position according to the formula:

[0021] Pactual = Pmeasured + β1(T - Tref) + β2(V - Vref) + β3×(Ppower - Pref)

[0022] where Tref represents the temperature reference parameter, Pref represents the pressure reference parameter, Vref represents the voltage reference parameter, and β1, β2, β3 represent correction coefficients.

[0023] In a preferred embodiment, in step 2, use a high-sampling-rate temperature sensor to sample the real-time temperature of the riveting head, and use a Butterworth high-pass filter to perform high-pass filtering on the temperature data to extract short-term fluctuation information according to the formula:

[0024]

[0025] Among them, Tmeasured(t) represents the collected temperature data, and Tfiltered(t) represents the short-term temperature fluctuation data after filtering;

[0026] The air velocity vair and flow direction dair data are collected by an air velocity sensor. The short-time Fourier transform is used to calculate the time-frequency spectra STFT(vair(t)) and STFT(dair(t)) of the air velocity vair and flow direction dair, and the K-means is used to identify different patterns of airflow disturbances and analyze local airflow disturbances;

[0027] The Kalman filter algorithm is used to jointly judge the temperature, air velocity data and high-resolution displacement data to correct the actual stop position of the riveting joint. According to the formula:

[0028] Pcorrected = K × Pactual + (1 - K) × Ptheory

[0029] Among them, K represents the weight for balancing the measured value and the predicted value in a noisy environment;

[0030] Calculate the deviation between the actual and the theoretical value. According to the formula: ΔP = Pcorrected - Ptheory, where ΔP represents the deviation, a positive value indicates that the actual position is ahead, and a negative value indicates insufficiency.

[0031] In a preferred embodiment, in step 3, a linear fitting model is used to construct the calculation formula for the initial compensation value A: A = K1 × (Ptheory - Pcorrected) + K0, where A is the initial compensation value, K1 is the compensation ratio coefficient, and K0 is the static compensation offset value;

[0032] Take A as the preset compensation amount and adjust the stroke control parameter Ptarget. According to the formula: Ptarget = Ptheory + A, where when under-riveting, the riveting-in stroke is increased; when over-riveting, the riveting-in stroke is decreased;

[0033] Use a laser displacement sensor to re-measure the new actual stop position Pactualnew after compensation; and calculate the new deviation value ΔPnew. According to the formula: ΔPnew = Ptheory - Pcorrectednew;

[0034] Calculate the error change before and after compensation according to the formula: Wc = (|ΔP| - |ΔPnew|) / |ΔP| × 100%, where Wc represents the error reduction rate. When Wc > 90%, the initial compensation value is considered valid; otherwise, the incremental update algorithm is used to optimize K1 and K0 according to the formula: K1new = K1 + α1 × (ΔPnew / ΔP), K0new = K0 + β1 × (ΔPnew / ΔP), where α1 and β1 are learning rate parameters.

[0035] In a preferred embodiment, in step 4, the online detection riveting device is used to continuously collect the actual stop position Pactual of each workpiece, and the current riveting stroke error B is calculated according to the formula:

[0036]

[0037] The stamping pressure Ppress is measured in real time by a pressure sensor, and the riveting force Friv actually acting on the workpiece is measured by a force sensor; and the stamping pressure deviation is calculated according to the formula:

[0038] ΔPpress = Ppress - Ppressref

[0039] where Ppressref represents the reference stamping pressure, ΔPress represents the stamping pressure deviation. Then, the additional stroke deviation Bpress caused by the pressure change is calculated by using the pressure - stroke experimental data fitting correction formula according to the formula:

[0040] Bpress = β × ΔPpress

[0041] where β represents the stamping pressure correction coefficient. Then, the correction calculation of the riveting stroke error is carried out according to the formula:

[0042] B′ = B - Bpress

[0043] where B′ represents the deviation after removing the influence of pressure;

[0044] The displacement data of the riveting head is measured in real time by an optical encoder, and the current speed V and acceleration a are calculated. Then, the riveting head speed deviation is calculated according to the formula: ΔV = V - Vref, where Vref represents the reference motion speed, ΔV represents the speed error; the riveting head acceleration deviation is calculated according to the formula: Δa = a - aref, where aref represents the reference acceleration, Δa represents the acceleration error; then, the additional stroke deviation Bmotion caused by the speed and acceleration changes is calculated by using the motion speed / acceleration - stroke error experimental data fitting correction formula according to the formula:

[0045] Bmotion = γ × ΔV + δ × Δa

[0046] Among them, γ is the speed influence coefficient, and δ is the acceleration influence coefficient; then, a correction calculation is performed on the deviation B′ according to the formula:

[0047] B′′ = B′ - Bmotion

[0048] Among them, B′′ represents the deviation after removing the influence of speed and acceleration.

[0049] In a preferred embodiment, in step 4, when the riveting device is in a stationary state, the wear amounts of the riveting head and the punch of the riveting device are measured by a laser rangefinder, the measurement data is recorded, and a timestamp is generated, including: wear amount, timestamp; and the built-in sensor of the riveting device is used to record the load information and usage duration of each use, as well as the timestamp;

[0050] Use GPS to measure the longitude, latitude coordinates, altitude, and timestamp of the current installation location of the riveting device; use an inclinometer to record the horizontal and vertical tilt angles and timestamp of the riveting device, and use a groundwater change monitor or a settlement gauge to collect the groundwater level or soil settlement value at the installation location of the riveting device and the timestamp;

[0051] Through regression analysis, calculate the wear correction amount ΔPwear according to the formula: ΔPwear = β2 × Fz + β3 × St, where Fz represents the load value of the riveting device, St represents the usage duration of the riveting device, and β2 and β3 represent correction coefficients; calculate the geological correction amount ΔPgeo according to the formula: ΔPgeo = α2 × Cj + α3 × Qx, where Cj represents the soil settlement value, Qx represents the tilt value, and α2 and α3 represent correction coefficients;

[0052] Calculate the new compensation value Anew according to the formula: Anew = Aprevious - α(B′′ - (ΔP_wear + ΔP_geo)), where α represents the deviation correction factor and 0 < α ≤ 1, and Aprevious represents the compensation value used for the previous workpiece;

[0053] Use the calculated compensation value Anew as the compensation input for the riveting stroke of the next workpiece, and monitor the correction effect of consecutive workpieces in real time, record the actual stop position and the deviation change curve after each adjustment, and form closed-loop feedback data.

[0054] The present invention discloses a method for automatically adjusting the riveting-in stroke of a non-nail rivet, which relates to the field of riveting technology and is used to solve the problem of riveting precision deviation. First, the theoretical stop position is determined through simulation analysis and experimental data, and a process reference is established based on the riveting process and material characteristics. Then, during the production process, the riveting stop position, ambient temperature, power supply voltage, and vibration data of the workpiece are collected in real time. The environmental interference is eliminated through a multi-factor correction method, the deviation is calculated and compared with the theoretical parameters to form dynamic compensation. To cope with long-term wear and changes in the installation environment, the present invention also calculates the wear correction amount and geological correction amount through regression analysis and neural network methods, combined with information such as wear data, load history, and geological environment, and dynamically adjusts the riveting stroke according to the corrected deviation. Finally, through real-time feedback and closed-loop control, the riveting process is optimized to achieve precise riveting positioning, improve production efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flowchart of the operation of a method for automatically adjusting the riveting-in stroke of a non-nail rivet according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiment

[0058] The present invention discloses a method for automatically adjusting the riveting-in stroke of a non-nail rivet, including:

[0059] Step 1: Preset the initial riveting-in stroke according to the design requirements and material characteristics, verify through simulation and experiments, calculate the optimal riveting-in stroke, and correct the theoretical stop position.

[0060] First, according to the product design and process requirements, including: product material characteristics, riveting process parameters, and structural stress analysis requirements, and based on past process experience, preset a preliminary riveting-in stroke range Pinit, and use the database matching method to retrieve the riveting stroke data of similar materials and structures to provide a reference initial value.

[0061] Furthermore, use CAD software such as SolidWorks and CATIA to draw the models of the workpiece, riveting head and support structure, import the models into FEA software, and divide them into high-precision meshes (mesh size <0.1 mm); meanwhile, use the bilinear hardening model to define the elastoplastic constitutive relationship based on the product material data, set the friction coefficient (the contact friction between the riveting head and the workpiece is 0.1 - 0.3). Then, fix the bottom workpiece, define the displacement constraint of the riveting head so that it only moves along the Z-axis, and apply a gradually increasing riveting pressure Ppress (5 - 20 kN). Finally, run the simulation, analyze the stress distribution, plastic deformation area and the structural strength after riveting, observe the deformation amount of the workpiece to ensure that there are no cracks or excessive springback after riveting, and through multiple rounds of simulation, find the optimal riveting stroke Pfea that satisfies uniform stress and sufficient load transfer.

[0062] Furthermore, to verify the simulation results, select several workpieces and rivet them into Pfea ± 0.1 mm for testing. Use a servo press for riveting, set the same punch speed and riveting pressure, record the riveting stroke Pmeasured of each workpiece, and eliminate the abnormal data that deviates from the mean value by ±3σ (standard deviation), and calculate the experimental stroke Pexp. According to the formula:

[0063]

[0064] where N is the number of valid experimental data points, and Pmeasured represents the actual stop position of the workpiece.

[0065] Furthermore, combine the simulation result Pfea and the experimental data Pexp to calculate the corrected theoretical stop position Ptheory according to the formula:

[0066] Ptheory = w1 × Pfea + w2 × Pexp

[0067] where w1 represents the weight of Pfea, and w2 represents the weight of Pexp.

[0068] Furthermore, calibrate the laser sensor according to the key parameters of the riveting stroke, punch speed and riveting pressure every 7 days, control the error within ±0.001 mm, and form a standardized process specification (SOP) to clarify the riveting equipment parameters, operation requirements and detection methods.

[0069] Step 2: Collect the actual stop position, temperature, power supply data, and preliminarily calculate the deviation;

[0070] First, install a laser displacement sensor on the fixed bracket at the riveting station, set the acquisition frequency to 1000 Hz, acquire the actual stop position Pmeasured of the workpiece, record the data and add a timestamp; fix a platinum resistance (RTD) temperature sensor on the metal frame of the riveting equipment, set the acquisition frequency of the temperature sensor to 10 Hz, and measure the ambient temperature data; install a power quality monitoring riveting device at the power input end of the riveting equipment, set the acquisition frequency of the power monitoring module to 10 Hz, synchronize with the temperature data, and record the current power voltage V and frequency f; install an IEPE vibration sensor at the punch part of the riveting equipment, set the acquisition frequency of the vibration sensor to 10 Hz, ensure that the data is synchronized with the temperature and power data, and record the current vibration RMS value V.

[0071] Furthermore, use an industrial data acquisition riveting device (DAQ) to associate each data source through timestamps to ensure that the actual stop position, temperature, power, and vibration data are all acquired at the same time point.

[0072] Furthermore, adopt a multi-factor correction method to eliminate the influence of temperature, power, and vibration on the actual stop position. According to the formula:

[0073] Pactual = Pmeasured + β1(T - Tref) + β2(V - Vref) + β3×(Ppower - Pref)

[0074] Where, Tref represents the temperature reference parameter, Pref represents the pressure reference parameter, Vref represents the voltage reference parameter, and β1, β2, β3 represent the correction coefficients.

[0075] It should be noted that in the production workshop, due to the sudden increase in local heat load during the riveting process, the temperature of the riveting head will rise instantaneously and sharply. At the same time, the start of the air conditioner in the workshop or the change in the position of the air outlet causes unstable local air flow, resulting in drastic fluctuations in the ambient temperature data. Simply eliminating the influence of temperature, power, and vibration cannot capture the instantaneous temperature fluctuations and air flow changes, resulting in large random deviations in the actual stop position data. Therefore, in this embodiment, when the riveting operation is started, a high-sampling-rate temperature sensor (500 Hz) is used to sample the temperature of the riveting head and the surrounding environment in real time and record the temperature of the riveting head. And a Butterworth high-pass filter (cutoff frequency 10 Hz) is used to perform high-pass filtering on the temperature data and extract the short-term fluctuation information. According to the formula:

[0076]

[0077] Where, Tmeasured(t) represents the acquired temperature data, and Tfiltered(t) represents the short-term temperature fluctuation data after filtering.

[0078] Furthermore, an air velocity sensor (200Hz) is installed in the riveting area to record the air velocity vair and the flow direction dair, and the short-time Fourier transform (STFT) is performed on the air velocity data to calculate the time-frequency spectra STFT(vair(t)) and STFT(dair(t)) of the air velocity vair and the flow direction dair. Through time-frequency analysis, it is judged whether the change in air velocity reaches the set disturbance threshold, and the K-means clustering analysis is used to analyze the flow velocity data to identify different patterns of airflow disturbance, capture the characteristics of local airflow changes, and analyze local airflow disturbance.

[0079] Furthermore, when the riveting head is about to stop, the filtered and feature-extracted temperature and airflow data are jointly judged with the high-resolution displacement data to eliminate the instantaneous noise caused by environmental disturbance, and the Kalman Filter algorithm is used to correct the actual stop position of the riveting head. According to the formula:

[0080] Pcorrected = K × Pactual + (1 - K) × Ptheory

[0081] where K represents the weight for balancing the measured value and the predicted value in the noise environment.

[0082] Furthermore, the deviation between the actual and the theoretical values is calculated according to the formula: ΔP = Pcorrected - Ptheory, where ΔP represents the deviation, a positive value indicates that the actual position is ahead (overshoot), and a negative value indicates insufficiency (under-riveting).

[0083] Step 3: Calculate the initial compensation value according to the deviation data and adjust the stroke parameters of the riveting equipment to verify the effectiveness of the compensation value.

[0084] First of all, according to the deviation data and the process parameters, a linear fitting model is used to construct the calculation formula for the initial compensation value: A = K1 × (Ptheory - Pcorrected) + K0, where A is the initial compensation value, K1 is the compensation ratio coefficient, and K0 is the static compensation offset.

[0085] Furthermore, using the above formula, the initial compensation value A for each workpiece is calculated according to the data of the first round of trial production. During the actual riveting process, A is used as the preset compensation amount, and in the PLC control program of the riveting machine, the stroke control parameter Ptarget is adjusted according to the formula: Ptarget = Ptheory + A. Where when under-riveting (ΔP > 0), the riveting stroke is increased; when over-riveting (ΔP < 0), the riveting stroke is decreased.

[0086] Further, after compensation, use a laser displacement sensor to re-measure the actual riveting stop position Pactualnew after compensation; and calculate the new deviation value ΔPnew according to the formula: ΔPnew = Ptheory - Pcorrectednew.

[0087] Further, calculate the error change before and after compensation according to the formula: error reduction rate = (|ΔP| - |ΔPnew|) / |ΔP|×100%. Among them, when the error reduction rate > 90%, the initial compensation value is considered valid; otherwise, the incremental update algorithm needs to be used to optimize K1 and K0 according to the formula: K1new = K1 + α1×(ΔPnew / ΔP), K0new = K0 + β1×(ΔPnew / ΔP), where α1 and β1 are learning rate parameters used to control the update speed of the compensation coefficient.

[0088] Step 4: Monitor the stop position in real time, collect key data, calculate and correct the stroke deviation caused by the wear of the riveting equipment and environmental changes, update the compensation value and perform feedback adjustment.

[0089] First, during the formal production process, use an online detection riveting equipment to continuously collect the actual stop position of each workpiece. Each measured actual position is compared with Ptheory, and the current deviation (denoted as B) is calculated according to the formula:

[0090]

[0091] Further, install a pressure sensor with an installation accuracy of 0.1 MPa and a sampling frequency of 100 Hz on the servo motor of the riveting equipment to measure the stamping pressure Ppress in real time. A force sensor (strain gauge) is fixed at the riveting head to measure the actual riveting force Friv acting on the workpiece. Establish the relationship between the stamping pressure Ppress and the riveting stroke error B according to the formula:

[0092] ΔPpress = Ppress - Ppressref

[0093] Among them, Ppressref represents the reference stamping pressure, and ΔPress represents the stamping pressure deviation. Then, use the pressure-stroke experimental data fitting correction formula to calculate the additional stroke deviation Bpress caused by the pressure change according to the formula:

[0094] Bpress = β×ΔPpress

[0095] Among them, β represents the stamping pressure correction coefficient. Then, perform correction calculation on the deviation according to the formula:

[0096] B′ = B - Bpress

[0097] Among them, B′ represents the deviation after removing the influence of pressure;

[0098] Furthermore, an optoelectronic encoder with an installation resolution of ±0.01 mm and a sampling frequency of 1000 Hz is installed on the main drive shaft or punch of the riveting equipment to measure the displacement data of the riveting head in real time, calculate the current speed V and acceleration a. An IEPE-type acceleration sensor with an acceleration range of ±50 g and a sampling frequency of 1000 Hz is fixed near the riveting punch to measure the instantaneous acceleration a. Then, calculate the speed deviation of the riveting head according to the formula: ΔV = V - Vref, where Vref represents the reference motion speed and ΔV represents the speed error; calculate the acceleration deviation of the riveting head according to the formula: Δa = a - aref, where aref represents the reference acceleration and Δa represents the acceleration error; then, use the experimental data fitting correction formula of motion speed / acceleration - stroke error to calculate the additional stroke deviation Bmotion caused by the changes in speed and acceleration according to the formula:

[0099] Bmotion = γ×ΔV + δ×Δa

[0100] Among them, γ is the speed influence coefficient and δ is the acceleration influence coefficient; then, perform a correction calculation on the deviation B′ according to the formula:

[0101] B′′ = B′ - Bmotion

[0102] Among them, B′′ represents the deviation after removing the influence of speed and acceleration;

[0103] It should be noted that the stamping pressure, motion speed, and acceleration parameters mainly reflect short-term dynamic changes and cannot reflect the cumulative influence of long-term wear of the riveting equipment and the installation foundation (such as geological conditions, foundation settlement) on the riveting accuracy, resulting in the compensation value being unable to accurately track the changes in the state of the riveting equipment. Therefore, in this embodiment, after each riveting is completed, ensure that the riveting equipment is in a stationary state, measure the wear amount of the key components (such as the riveting head, punch) of the riveting equipment through a laser rangefinder or an optical detector, record the measurement data, and generate a time stamp, the content including: wear amount, time stamp; and record the load information and usage duration of each use through the built-in sensor of the riveting equipment, and the recorded data includes: load history, usage duration, time stamp.

[0104] Furthermore, before and after each riveting, use a GPS to measure the longitude, latitude coordinates, altitude, and time stamp of the current installation position of the riveting equipment, ensure that each measurement is synchronized with the data acquisition time. After each riveting is completed, use an inclinometer to record the horizontal and vertical tilt angles and time stamp of the riveting equipment, and use a groundwater change monitor or a settlement gauge to regularly collect the underground water level or soil settlement value and time stamp at the installation location of the riveting equipment and save them in the form of periodically recorded data.

[0105] Further, through regression analysis, the relationship between the wear amount, the load of the riveting equipment, and the usage time is quantified, and the wear correction amount ΔPwear is calculated according to the formula: ΔPwear = β2×Fz + β3×St, where Fz represents the load value of the riveting equipment, St represents the usage duration of the riveting equipment, and β2 and β3 represent correction coefficients; using the tilt data and soil settlement data, the influence of the installation environment on the riveting equipment is evaluated, and the geological correction amount ΔPgeo is calculated according to the formula: ΔPgeo = α2×Cj + α3×Qx, where Cj represents the soil settlement value, Qx represents the tilt value, and α2 and α3 represent correction coefficients;

[0106] Further, according to the idea of "Result A - B", a new compensation value Anew is calculated in combination with the corrected deviation B′′ according to the formula: Anew = Aprevious - α(B′′ - (ΔP_wear + ΔP_geo)), where α represents the deviation correction factor and 0 < α ≤ 1, and Aprevious represents the compensation value used in the previous workpiece or the current cycle.

[0107] Further, the calculated compensation value Anew is used as the compensation input for the riveting stroke of the next workpiece, and the correction effect of consecutive workpieces is monitored in real time, and the actual stop position and the deviation change curve after each adjustment are recorded to form closed-loop feedback data.

[0108] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0109] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0110] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0111] In addition, in each embodiment of the present application, the functional modules can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0112] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

[0113] Finally: The above description is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for automatically adjusting the riveting stroke of a nail-free rivet, characterized in that; It includes the following steps: Step 1: Preset the initial riveting stroke, calculate the optimal riveting stroke, and correct the theoretical stop position; Step 2: Collect the actual stop position, temperature, power supply, and air flow rate data, and preliminarily calculate the deviation; Step 3: Calculate the initial compensation value, adjust the stroke parameters of the riveting equipment, and verify the effectiveness of the compensation value; Step 4: Calculate and correct the stroke deviation caused by the wear of the riveting equipment and environmental changes, update the compensation value, and perform feedback adjustment; Preset the initial riveting stroke range Pinit, use the CAD software SolidWorks to draw the models of the workpiece, riveting head, and support structure, import them into the FEA software, divide the high-precision mesh, and at the same time, define the displacement constraints of the riveting head and apply the riveting pressure Ppress. Finally, through multiple rounds of simulation, find the optimal riveting stroke Pfea; Select several workpieces to be riveted with the optimal riveting stroke Pfea±0.1mm for testing, use a servo press for riveting, set the same punch speed and riveting pressure, and eliminate the abnormal data deviating from the mean ±3σ standard deviation, calculate the experimental stroke Pexp, according to the formula: ； where N is the number of data points after eliminating the abnormal data, and Pmeasured represents the actual stop position of the workpiece; Calculate the corrected theoretical stop position Ptheory, according to the formula: Ptheory=w1×Pfea+w2×Pexp; where w1 represents the weight of the optimal riveting stroke Pfea, and w2 represents the weight of the experimental stroke Pexp.

2. The automatic adjustment method for the riveting stroke of nail-free riveting according to claim 1, wherein: In Step 2, use a laser displacement sensor to collect the actual stop position Pmeasured of the workpiece, use a platinum resistance temperature sensor to collect the ambient temperature data, use a power quality monitor of the riveting equipment to collect the current power supply voltage V and frequency f, and use an IEPE vibration sensor to collect the current vibration value V; Use the multi-factor correction method to eliminate the influence of temperature, power supply, and vibration on the actual stop position, according to the formula: Pactual=Pmeasured+β1(T-Tref)+β2(V-Vref)+β3×(Ppower-Pref) where Tref represents the temperature reference parameter, Vref represents the voltage reference parameter, and β1, β2, β3 represent the correction coefficients.

3. A method for automatically adjusting the riveting stroke of nail-free riveting according to claim 2, characterized in that; In Step 2, use a high-sampling-rate temperature sensor to sample the real-time temperature of the riveting head, and use a Butterworth high-pass filter to perform high-pass filtering on the temperature data to extract the short-term fluctuation information, according to the formula: ； where Tmeasured(t) represents the collected temperature data, and Tfiltered(t) represents the short-term temperature fluctuation data after filtering; Use an air flow rate sensor to collect the air flow rate vair and flow direction dair data, use the short-time Fourier transform to calculate the time-frequency spectra STFT(vair(t)) and STFT(dair(t)) of the air flow rate vair and flow direction dair, and use K-means to identify different modes of air flow disturbance and analyze the local air flow disturbance; The Kalman filter algorithm is used to jointly judge the temperature, air flow velocity data and high-resolution displacement data, and correct the actual stop position of the riveting head according to the formula: Pcorrected = K × Pactual + (1 - K) × Ptheory; where K represents the weight for balancing the measured value and the predicted value in the noise environment; Calculate the deviation between the actual and the theoretical values according to the formula: ΔP = Pcorrected - Ptheory, where ΔP represents the deviation, a positive value indicates that the actual position is ahead, and a negative value indicates insufficiency.

4. A method for automatically adjusting the riveting travel of nail-free riveting according to claim 3, characterized in that: In step 3, a linear fitting model is used to construct the calculation formula for the initial compensation value A: A = K1 × (Ptheory - Pcorrected) + K0, where A is the initial compensation value, K1 is the compensation ratio coefficient, and K0 is the static compensation offset value; Take A as the preset compensation amount and adjust the stroke control parameter Ptarget according to the formula: Ptarget = Ptheory + A, where when under-riveting, the riveting-in stroke is increased; when over-riveting, the riveting-in stroke is decreased; Use a laser displacement sensor to re-measure the new actual stop position Pactualnew after compensation; and calculate the new deviation value ΔPnew according to the formula: ΔPnew = Ptheory - Pcorrectednew; Calculate the error change before and after compensation according to the formula: Wc = (|ΔP| - |ΔPnew|) / |ΔP| × 100%, where Wc represents the error reduction rate. When Wc > 90%, the initial compensation value is considered effective; otherwise, use the incremental update algorithm to optimize K1 and K0 according to the formula: K1new = K1 + α1 × (ΔPnew / ΔP), K0new = K0 + β1 × (ΔPnew / ΔP), where α1 and β1 are learning rate parameters.

5. A method for automatically adjusting the riveting stroke of a nail-free rivet according to claim 4, characterized in that: In step 4, use the on-line detection riveting equipment to continuously collect the actual stop position Pactual of each workpiece, and calculate the current riveting-in stroke error B according to the formula: ； Use a pressure sensor to measure the stamping pressure Ppress in real time, and use a force sensor to measure the riveting force Friv actually acting on the workpiece; and calculate the stamping pressure deviation according to the formula: ΔPpress = Ppress - Ppressref; where Ppressref represents the reference stamping pressure, ΔPress represents the stamping pressure deviation, and then use the pressure-stroke experimental data fitting correction formula to calculate the additional stroke deviation Bpress caused by the pressure change according to the formula: Bpress = β × ΔPpress; where β represents the stamping pressure correction coefficient, and then correct and calculate the riveting-in stroke error according to the formula: B′ = B - Bpress; where B′ represents the deviation after removing the influence of the pressure; The displacement data of the riveting joint is measured in real time using an optoelectronic encoder, and the current speed V and acceleration a are calculated. Then, the speed deviation of the riveting joint is calculated according to the formula: ΔV = V - Vref, where Vref represents the reference motion speed and ΔV represents the speed error; the acceleration deviation of the riveting joint is calculated according to the formula: Δa = a - aref, where aref represents the reference acceleration and Δa represents the acceleration error; then, the additional stroke deviation Bmotion caused by changes in speed and acceleration is calculated using the experimental data fitting correction formula of motion speed / acceleration - stroke error according to the formula: Bmotion = γ×ΔV + δ×Δa; where γ is the speed influence coefficient and δ is the acceleration influence coefficient; then the deviation B′ is corrected and calculated according to the formula: B′′ = B′ - Bmotion; where B′′ represents the deviation after removing the influence of speed and acceleration.

6. A method for automatically adjusting the riveting stroke of a nail-free rivet according to claim 5, characterized in that: In step 4, when the riveting equipment is in a stationary state, the wear amounts of the riveting joint and the punch of the riveting equipment are measured by a laser rangefinder, the measurement data is recorded, and a timestamp is generated, including: wear amount, timestamp; and the load information and usage duration of each use, as well as the timestamp, are recorded using the built-in sensors of the riveting equipment; The GPS is used to measure the longitude and latitude coordinates, altitude, and timestamp of the current installation position of the riveting equipment; the inclinometer is used to record the horizontal and vertical tilt angles and timestamp of the riveting equipment, and the groundwater change monitor or settlement gauge is used to collect the groundwater level or soil settlement value at the installation location of the riveting equipment and the timestamp; Through regression analysis, the wear correction amount ΔPwear is calculated according to the formula: ΔPwear = β2×Fz + β3×St, where Fz represents the load value of the riveting equipment, St represents the usage duration of the riveting equipment, and β2 and β3 represent the correction coefficients; the geological correction amount ΔPgeo is calculated according to the formula: ΔPgeo = α2×Cj + α3×Qx, where Cj represents the soil settlement value, Qx represents the tilt value, and α2 and α3 represent the correction coefficients; The new compensation value Anew is calculated according to the formula: Anew = Aprevious - α(B′′ - (ΔP_wear + ΔP_geo)), where α represents the deviation correction factor and 0 < α ≤ 1, and Aprevious represents the compensation value used for the previous workpiece; The calculated compensation value Anew is used as the compensation input for the riveting stroke of the next workpiece, and the correction effect of consecutive workpieces is monitored in real time, and the actual stop position and deviation change curve after each adjustment are recorded to form closed-loop feedback data.

Images