Air spring buckling process compensation control method and device and storage medium

Abstract

This invention discloses a compensation control method, device, and storage medium for the air spring crimping process. The method includes: acquiring a gap measurement value and comparing it with a preset standard gap value to form a gap deviation vector; using a gap deviation-displacement compensation mapping model to determine the differentiated target displacement of each channel after compensation; sending the compensated target displacement vector to the crimping control module to execute the crimping action and collect the actual final displacement value; determining the actual radial compression amount of each channel, and evaluating the product roundness based on the consistency of the actual radial compression amounts of each channel. This invention actively converts the gap deviation vector obtained by visual inspection before crimping into a differentiated displacement compensation amount for each channel, eliminating the influence of gap unevenness on the roundness after crimping through feedforward compensation, effectively improving the product roundness control accuracy and effective output rate.

Description

Technical Field

[0001] This invention relates to the field of automotive parts assembly process control technology, and in particular to a method, device and storage medium for compensating and controlling the air spring clamping process. Background Technology

[0002] Commercial vehicle seat air spring assemblies are key components ensuring driving comfort and safety, and the crimping process in their production directly determines the product's airtightness and service life. The crimping process uses a multi-lobed crimping mechanism to press a metal crimping ring with radial contraction force onto the mating surface between the airbag and the main body, ensuring a reliable seal between the airbag's rubber layer and the mating surface.

[0003] In the production practice of air spring crimping, poor roundness after crimping is one of the main quality defects affecting sealing reliability and product lifespan. Regarding roundness control, the industry has long developed the following technical understandings and solutions.

[0004] Industry experts generally believe that the main cause of poor roundness after crimping stems from the precision of the crimping mechanism itself, including insufficient synchronization of the clamping segments, uneven force distribution between segments due to wear, and pressure fluctuations in the hydraulic system. Based on this understanding, mainstream improvement efforts focus on increasing the manufacturing precision and maintenance frequency of the crimping mechanism, optimizing the pressure stability of the hydraulic circuit, and screening out products manufactured under abnormal conditions through force monitoring. These measures have improved roundness control to some extent, but even after the precision of the crimping mechanism has reached a high level, the roundness defect rate still faces a bottleneck that cannot be further reduced. Furthermore, the occurrence of defective products is random, making it difficult to pinpoint a definitive cause from the perspective of the mechanism's condition.

[0005] In recent years, with the widespread adoption of machine vision technology in manufacturing, some production lines have introduced vision inspection systems to check the assembly status before crimping. The vision system can detect the uniformity of the gap between the crimping ring and the main component, stopping crimping when the gap distribution deviates significantly from the standard to avoid producing obviously defective products. However, the functional positioning of this application has always been limited to determining the pass / fail threshold of the assembly status; that is, if it fails, the machine is stopped and the material is returned; if it passes, it is released to the subsequent standard crimping process. The quantitative data on the gap distribution obtained by vision inspection is discarded after the release judgment is completed and is not further utilized.

[0006] The industry's practice of using visual inspection data only for threshold determination without further utilization is based on the following technical understanding: First, the threshold determination has already screened out products with serious gap deviations, and the gap deviations of the released products are considered to be within a range that has no significant impact on the crimping result; Second, the crimping mechanism applies uniform control parameters to each channel, which is considered sufficient to ensure that the released products meet the roundness requirements after crimping. That is, as long as the material feeding state is within the tolerance, the standard process parameters can guarantee the quality of the output.

[0007] Specifically, existing crimping quality control methods mainly include the following types. The first method only monitors the final value of the combined force of all flaps and sets a acceptable force range for judgment. Its drawback is that even if the total force value is acceptable, the force between each flap may be severely uneven, but this unevenness cannot be reflected after the total force value is averaged. The second method independently collects force and displacement data for each flap and compares it with the acceptable envelope. Although it can identify single-channel anomalies, its envelope setting faces a dilemma: a wide envelope cannot intercept roundness defects, while a narrow envelope leads to frequent misjudgments of normal products. Furthermore, this method is only passive detection rather than active correction; when a defect is detected, the product has already been crimped and must be scrapped. The third method uses a vision system to detect the uniformity of the gap before crimping. If it is unacceptable, crimping is stopped; if it is acceptable, it is executed according to standard parameters. As mentioned earlier, its function is limited to threshold judgment.

[0008] In production practice, a phenomenon has been observed: even with high-precision crimping mechanisms and reasonable visual inspection thresholds, a significant proportion of products still fail to meet roundness standards after crimping. Offline analysis of these defective products revealed that their pre-crimping gap deviation values ​​are typically within the visual inspection release threshold, falling into the gray-scale range of products that pass inspection but have poor roundness. The industry's conventional approach for such products is to tighten the visual inspection release threshold to reduce the gray-scale range. However, this leads to a large number of products that could have passed being mistakenly rejected and returned, significantly reducing the effective output rate. Furthermore, even tightening the threshold cannot reduce the roundness defect rate to an ideal level.

[0009] More importantly, analysis of the actual crimping displacement data of each channel of the defective product showed that the displacement control accuracy of each channel was good (i.e., each channel reached the set target displacement value), but the roundness after crimping was still unqualified. This phenomenon further reinforced the erroneous attribution direction that the roundness problem was unrelated to the state of the feed gap, but related to the precision of the mechanism or the batch of materials, because on the surface, the feed gap was within the acceptable range and the displacement control was accurate, so the roundness defect could not be attributed to the gap deviation.

[0010] The aforementioned situation constitutes a long-standing technical dilemma in this field: the roundness control accuracy is at a bottleneck, but the cause is unclear; the utilization of visual inspection data remains at the binary judgment level; and the crimping execution parameters are uniformly set for all released products without differentiation. The industry has failed to establish a framework for the correlation between quantitative data on the product's state before crimping and the execution parameters of the crimping process, and therefore has not developed a technical approach to proactively use pre-inspection data to adjust subsequent process parameters. For products like air springs that contain elastic rubber materials, if the state deviation before crimping is not actively compensated, it will directly translate into roundness deviation after crimping, leading to problems such as insufficient roundness control accuracy and decreased effective yield. Summary of the Invention

[0011] In view of the above problems, the present invention provides a method, device and storage medium for compensation control of air spring crimping process, which can convert the circumferential gap deviation vector obtained by visual inspection before crimping into the differential crimping displacement compensation amount of each channel. By actively eliminating the influence of uneven gap on the roundness after crimping through feedforward compensation, the roundness control accuracy is improved and the output rate is increased.

[0012] In some embodiments, a compensation control method for the air spring clamping process is provided, comprising: acquiring gap measurement values ​​at multiple channel positions in the circumferential direction between the clamping ring and the main body before clamping; determining the gap deviation value of each channel based on the gap measurement values ​​and a preset standard gap value, forming a gap deviation vector; determining the target displacement after compensation for each channel based on the gap deviation vector using a gap deviation-displacement compensation mapping model, forming a target displacement vector after compensation; sending the target displacement vector after compensation to the clamping control module, performing the clamping action with the independent target displacement after compensation for each channel as the displacement control endpoint, and collecting the actual final displacement value of each channel; determining the actual radial compression amount of each channel based on the actual final displacement value of each channel and the gap deviation vector, and evaluating the roundness of the product based on the consistency of the actual radial compression amount of each channel.

[0013] In this embodiment, by quantizing the continuous gap deviation vector obtained by visual inspection before buckling into the differentiated compensation target value of each channel, the pre-detection information is elevated from threshold judgment to feedforward control basis, so that the roundness deviation caused by uneven gap is actively eliminated during buckling, rather than being passively discovered and rejected after buckling.

[0014] In some embodiments, the gap deviation-displacement compensation mapping model is a piecewise compensation model, and the compensation coefficients include positive compensation coefficients and negative compensation coefficients. When the gap deviation value of the i-th channel is greater than or equal to zero, the compensated target displacement of the i-th channel is determined to be the sum of the standard target displacement value, the positive compensation coefficient, and the gap deviation value of that channel; when the gap deviation value of the i-th channel is less than zero, the compensated target displacement of the i-th channel is determined to be the sum of the standard target displacement value, the negative compensation coefficient, and the gap deviation value of that channel. The positive compensation coefficient is greater than 1, and the negative compensation coefficient is less than or equal to 1.

[0015] In this embodiment, due to the elastic deformation and flow effect of the elastic airbag material during the buckling process, the displacement control system has a systematic tracking deviation related to the rubber stiffness characteristics under different load conditions. The setting of the positive and negative differential compensation coefficients is determined by calibration test to minimize the roundness index, thereby effectively compensating for the influence of this systematic deviation on the roundness after buckling.

[0016] In some embodiments, before determining the compensated target displacement of each channel based on the gap deviation vector, a validity check of the gap deviation vector is further included. The validity check includes calculating the range and maximum offset of the gap deviation values ​​for all channels. If the range exceeds a preset upper limit threshold or the maximum offset exceeds a preset maximum offset threshold, the gap deviation vector is determined to have failed the validity check, the clamping is stopped, and an alarm is triggered.

[0017] In this embodiment, the validity verification mechanism can identify products with severely abnormal gap distribution before compensation calculation, avoiding the execution of potentially invalid or even harmful compensation clipping on such products, thus ensuring the safety and reliability of compensation control.

[0018] In some embodiments, the method further includes limiting the compensation amount of each channel. If the absolute value of the compensation amount of any channel exceeds the preset maximum allowable compensation amount, it is determined that the gap deviation of the product exceeds the compensable range, the clamping is stopped, and an alarm is triggered. At the same time, the consistency of the target displacement vector after compensation is checked. If the range of the target displacement of each channel exceeds the preset upper limit of the range, it is determined to be in an uncompensable state.

[0019] In this embodiment, a dual protection mechanism of compensation amount limit and consistency verification is used to prevent excessive compensation amount from causing local stress exceeding the limit or rubber damage, and to clearly distinguish between normal products within the compensation range and abnormal products that require manual intervention.

[0020] In some embodiments, evaluating product roundness further includes performing multi-layer judgment logic: single-channel positioning judgment, force-displacement curve envelope judgment, roundness evaluation based on the consistency of actual radial compression, and comprehensive judgment. The qualified envelope of the force-displacement curve is adjusted accordingly based on the compensation amount of each channel.

[0021] In this embodiment, the multi-layer judgment logic comprehensively evaluates the crimping quality from different dimensions. In particular, the design of the force-displacement curve envelope shifting with the compensation amount solves the problem of the failure of the traditional fixed envelope judgment method under the adaptive compensation mode.

[0022] In some embodiments, the method further includes performing a channel group balance statistical analysis on the compression deviation values ​​of each channel of a series of qualified products. When the difference between the average compression deviation of the positive compensation channel group and the negative compensation channel group shows a systematic shift, the compensation coefficient is iteratively corrected using an incremental adjustment method, and the corrected coefficient is subject to reasonable range constraints.

[0023] In this embodiment, the compensation coefficient is iteratively corrected by directly using the channel group balance of compression deviation as the feedback index, so that the feedback quantity is precisely aligned with the roundness optimization target. That is, under the typical working condition where the initial calibration has minimized the intra-group dispersion and the production drift is mainly manifested as the inter-group common mode shift, when the balance index is maintained at the calibration reference value, the optimal roundness state established by the initial calibration is maintained. This allows the compensation model to automatically adapt to slow common mode drift factors such as batch fluctuations in airbag rubber hardness and gradual changes in characteristics caused by wear of the buckling mechanism, and maintain the compensation accuracy in a stable convergence manner over a long period of time.

[0024] According to a second aspect of the present invention, a multi-channel adaptive compensation control device for the air spring clamping process is provided, comprising a gap detection module, a compensation calculation module, a clamping execution control module, and a roundness evaluation module. The modules work together to realize a complete adaptive control process from gap detection to compensation calculation to clamping execution to roundness verification.

[0025] According to a third aspect of the present invention, an air spring clamping system is provided, comprising a vision inspection system, a multi-lobed clamping mechanism, and a controller, the controller being configured to execute the above-described multi-channel adaptive compensation control method.

[0026] According to a fourth aspect of the present invention, an electronic device is provided, including a memory and a processor coupled to the memory, the processor being configured to execute the above-described multi-channel adaptive compensation control method based on instructions stored in the memory.

[0027] According to a fifth aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described multi-channel adaptive compensation control method.

[0028] The air spring clamping process compensation control method, apparatus, electronic device, system, and storage medium provided by this invention have at least the following beneficial effects: 1. Significantly Improved Roundness Control Accuracy: By converting the pre-crimping gap deviation vector into differentiated displacement compensation amounts for each channel, a feedforward adjustment relationship is established between the pre-crimping gap deviation state and the crimping execution parameters, proactively eliminating roundness deviations caused by uneven gaps at the source. Compared to the existing technology that only uses visual inspection results for binary pass / fail judgment, this invention quantifies the continuous gap deviation vector and transforms it into a feedforward control basis, improving the information utilization level. This technical approach of establishing a quantitative mapping relationship between pre-inspection data and subsequent execution parameters overcomes the technical bias of existing technologies where detection and execution are mutually isolated.

[0029] 2. Improved effective output: For products in the gap deviation range that is not enough to trigger a shutdown but is enough to affect the roundness grayscale, the existing method can only find the non-conformity after crimping and reject them. This method can make the roundness of these products meet the standard after crimping through compensation, reducing the scrap caused by poor roundness. 3. Balancing sealing reliability and rubber protection: Differentiated displacement compensation is performed by using positive and negative differential compensation coefficients determined by calibration tests, so that the actual radial compression in each direction tends to be consistent after compensation. This avoids both local insufficient compression leading to seal leakage and local excessive compression leading to rubber damage. This segmented compensation model can effectively compensate for the displacement control system linear deviation related to the stiffness characteristics of elastic materials. 4. The compensation model has self-learning capabilities: By using the balance of compression deviation between positive and negative channel groups as a feedback index to continuously iterate and correct the compensation coefficients, under the typical working condition where the initial calibration has optimized the dispersion within the group and the production drift is the common mode offset between groups, this feedback index provides the correct correction direction when the roundness optimum is equal to the calibration reference value or deviates from the reference value, so that the feedback mechanism has a stable equilibrium point and converges stably under the learning rate value. The model can automatically adapt to slow-changing factors such as batch fluctuations in airbag rubber hardness and wear of the buckling mechanism, and maintain compensation accuracy for a long time without frequent manual calibration. Attached Figure Description

[0030] Figure 1 The overall control flowchart of the air spring clamping process compensation control method provided in the embodiments of the present invention is shown below; Figure 2 This is a schematic diagram illustrating the principle of visual inspection of gap distribution before buckling provided in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the calculation principle of the compensation model provided in the embodiments of the present invention, showing the mapping relationship between the gap deviation vector and the compensation target displacement vector; Figure 4A schematic diagram of the stiffness characteristic curves of the elastic airbag material in different compression sections provided in the embodiments of the present invention; Figure 5 This is a schematic diagram of the roundness evaluation method after buckling provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the feedback control logic for iterative correction of compensation model coefficients provided in an embodiment of the present invention; Figure 7 This is a schematic diagram showing the comparison of the product roundness index distribution before and after compensation, provided in an embodiment of the present invention. Figure 8 This is a functional block diagram of the multi-channel adaptive compensation control device provided in an embodiment of the present invention; Figure 9 An architectural block diagram of an electronic device provided in an embodiment of the present invention; Figure 10 This is a schematic diagram of the overall architecture and signal flow of the air spring clamping system provided in an embodiment of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0032] It should be noted that the ordinal numbers such as "first" and "second" described in the embodiments of the present invention are for distinguishing purposes only and should not be construed as a specific order restriction. For example, "first channel" and "second channel" are only used to distinguish different channels and do not represent a specific temporal or priority relationship between the channels.

[0033] In the embodiments of this invention, "and / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A existing alone, A and B existing simultaneously, and B existing alone.

[0034] Unless otherwise specified, the data transmission and signal sending operations involved in the embodiments of this invention refer to the electrical signal transmission between functional modules within an electronic device or system via data buses, communication interfaces, or signal lines. Terms such as "sending," "acquiring," and "feedback" between modules refer to the transmission process of data or control signals at the electrical level.

[0035] The background technology involved in the technical solution of this invention will be introduced below.

[0036] In the clamping process of commercial vehicle seat air spring assemblies, the clamping mechanism typically employs a multi-lobed structure. Each lobe is evenly distributed circumferentially and simultaneously contracts radially towards the center under driving force, uniformly pressing the clamping ring against the mating surface of the airbag and the main body. Each lobe corresponds to an independent clamping channel, and each channel is equipped with an independent displacement sensor to monitor its radial displacement in real time. Ideally, when the initial position of the clamping ring is perfectly concentric with the main body and the airbag is evenly distributed circumferentially, each channel performs clamping according to the same standard target displacement. The force and displacement of each channel are consistent, ensuring the roundness of the product after clamping meets requirements.

[0037] However, in actual production, due to factors such as local wrinkles or offsets in the elastic rubber material of the airbag, the ellipticity tolerance of the thin-walled metal parts of the clamping ring, and the repeatability error of the positioning fixture of the main body, the annular gap between the clamping ring and the main body is often unevenly distributed before clamping. If this uneven distribution is not actively compensated, it will be directly transmitted as a roundness deviation during the clamping process. During the clamping process, due to the elastic deformation and flow of the elastic airbag rubber under pressure, the displacement control system faces different load characteristics under different gap conditions. That is, the rubber of the channel with a larger gap is in the low compression section with lower initial stiffness, while the rubber of the channel with a smaller gap is in the high compression section with higher initial stiffness. This difference in load stiffness causes the displacement control system to produce a directional systematic tracking deviation for each channel. That is, the channel with a larger gap tends to be under-compressed, and the channel with a smaller gap tends to be over-compressed. Simply compensating by equal amount according to the gap deviation value cannot completely eliminate this systematic deviation. The core technical idea of ​​this invention is to establish a closed-loop control system of feedforward compensation-clipping execution-feedback verification-model iteration, which transforms the gap deviation vector detected before clamping into differentiated control parameters of each channel after calibration and optimization, thereby eliminating the influence of uneven gap and the resulting systematic control deviation on roundness.

[0038] The air spring clamping process compensation control method provided by the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0039] like Figure 1 As shown, the air spring clamping process compensation control method provided in this embodiment of the invention includes the following steps: Step S101: Obtain the gap measurement values ​​at multiple channel positions in the circumferential direction between the clamping ring and the main body before clamping. Based on the gap measurement values ​​and the preset standard gap values, determine the gap deviation value of each channel to form a gap deviation vector.

[0040] like Figure 2As shown, after the product arrives at the crimping station and completes positioning, the vision inspection system acquires a high-resolution image of the crimping area from above the product. The system uses a circular fitting algorithm to identify and fit the outer edge circle of the crimping ring and the outer edge circle of the main body, obtaining the center coordinates and radius parameters of both. Then, according to the angular position of each crimping channel, the system extracts the radial distance between the outer edge of the crimping ring and the outer edge of the main body in each angular direction, as the gap value in that direction. Since the radial wall thickness of the crimping ring is a known constant, this wall thickness constant is already included in the preset standard gap value G0 when measuring from the outer edge, and does not affect the gap deviation value ΔG. i The calculation results.

[0041] In some embodiments, the clamping mechanism has N independent control channels evenly distributed along the circumference. Taking N=12 as an example, each channel is numbered i, i=1,2,...,12, corresponding to angular positions of 0°, 30°, 60° up to 330°. The standard gap value of each channel is G0, defined by the product formulation parameters.

[0042] The formula for calculating the channel gap deviation is as follows: ΔG i =G i -G0 Among them, G i G is the measured gap value of the i-th channel, G0 is the standard gap value, and ΔG i Let be the gap deviation value of the i-th channel. The gap deviation values ​​of all channels constitute the gap deviation vector ΔG=[ΔG1,ΔG2,...,ΔG...]. 12 ].

[0043] When ΔG i When ΔG > 0, it indicates that the gap at the i-th channel position is greater than the standard value, meaning the distance between the clamping ring and the main body is too large in that direction. If clamping is still performed according to the standard displacement, the final radial compression in that direction will be too small, resulting in insufficient clamping. i When the value is less than 0, it means that the gap at the i-th channel position is less than the standard value. The clamping ring is too close in this direction. If it is still clamped according to the standard displacement, the final radial compression in this direction will be too large and the clamping will be too tight.

[0044] For example, assuming a product's standard gap value G0 = 2.50mm, the measured gap values ​​obtained through visual inspection for 12 channels are as follows: G1 = 2.65mm, G2 = 2.58mm, G3 = 2.52mm, G4 = 2.45mm, G5 = 2.38mm, G6 = 2.42mm, G7 = 2.48mm, G8 = 2.55mm, G9 = 2.62mm, G... 10 =2.60mm, G 11 =2.55mm, G 12=2.58mm. Therefore, the gap deviations for each channel are: ΔG1=+0.15mm, ΔG2=+0.08mm, ΔG3=+0.02mm, ΔG4=-0.05mm, ΔG5=-0.12mm, ΔG6=-0.08mm, ΔG7=-0.02mm, ΔG8=+0.05mm, ΔG9=+0.12mm, ΔG... 10 =+0.10mm、ΔG 11 =+0.05mm, ΔG 12 =+0.08mm.

[0045] In this embodiment, a visual inspection system is used to quantitatively measure the gap before crimping, rather than simply making a pass / fail threshold judgment. This obtains quantitative data on the circumferential gap distribution, providing accurate input information for subsequent differential compensation calculations. This step is the basis for upgrading the pre-detection information from the original binary judgment to the basis of feedforward control.

[0046] Step S102: Verify the validity of the gap deviation vector.

[0047] In some embodiments, before proceeding with compensation calculations, the system first verifies the validity of the gap deviation vector to ensure that the gap distribution of the product is within the effective working range of adaptive compensation.

[0048] The validity verification includes the calculation and judgment of two indicators. The first is the range verification: calculating the range value R = max(ΔG) of the gap deviation of all channels. i )-min(ΔG i The second item is the maximum offset check: calculate the maximum value M among the absolute values ​​of the gap deviation values ​​of each channel. max =max(|ΔG i |).

[0049] When the range value R exceeds the preset upper limit threshold R max In the case of, or at the maximum offset M max If the offset exceeds the preset maximum threshold, the system determines that the product's gap distribution is severely abnormal and unsuitable for adaptive compensation clamping. Clamping is stopped and an alarm is triggered, requiring manual intervention. The system only proceeds to the subsequent compensation calculation stage if both checks pass.

[0050] Continuing with the numerical example above, the range R = max(ΔG) i )-min(ΔG i =0.15 - (-0.12) = 0.27 mm. Assume a preset upper limit threshold R. max =0.40mm, the maximum offset threshold is 0.25mm, then R=0.27mm <R max =0.40mm, Mmax =0.15mm<0.25mm, both checks passed, and the system proceeds to the compensation calculation stage.

[0051] In this embodiment, by placing validity verification before compensation calculation, products with severely abnormal gap distribution can be screened out at an early stage, avoiding potentially invalid or even harmful compensation buckling operations on such products, thereby ensuring the safety of compensation control, and at the same time establishing an effective input data quality guarantee for subsequent compensation calculation.

[0052] Step S103: Based on the gap deviation vector, use the gap deviation-displacement compensation mapping model to determine the compensated target displacement of each channel and form the compensated target displacement vector.

[0053] like Figure 3 As shown, this step is the core calculation process. Its technical essence lies in establishing a mapping model from the gap deviation vector to the displacement compensation vector, which quantitatively converts the state deviation of each channel detected before crimping into the differentiated displacement commands that each channel needs to execute.

[0054] In some embodiments, the gap deviation-displacement compensation mapping model is a piecewise compensation model. The standard crimping target displacement value for each channel is D0, defined by the product formulation parameters. The compensation coefficients include the positive compensation coefficient Kc. + and negative compensation coefficient Kc - .

[0055] like Figure 4 As shown, the airbag rubber is an elastic material, and its stress-strain relationship exhibits typical nonlinear characteristics. When the buckle compresses the rubber, the elastic deformation and flow effect of the rubber affect the actual tracking performance of the displacement control system. Specifically: for channels with larger gaps, the rubber is in the lower section of the compression stroke, with lower initial stiffness. Under this low-stiffness load, the displacement control system tends to produce a systematic deviation of under-tracking (i.e., the actual effective compression amount is less than the commanded amount). This deviation means that even if the compensation coefficient is 1, the final radial compression amount in this direction is still too small. For channels with smaller gaps, the rubber is in the higher section of the compression stroke, with higher initial stiffness. Under this high-stiffness load, the tracking deviation of the displacement control system is opposite in direction and smaller in amplitude.

[0056] Therefore, the compensation coefficient Kc needs to be determined through system calibration to minimize the roundness index Rc after crimping. Statistical analysis of the calibration tests revealed that channels with larger gaps require a compensation amount slightly greater than the gap deviation value to offset the systematic under-tracking deviation in the low-stiffness section. Therefore, the positive compensation coefficient Kc... + >1; The systematic deviation of the channel with a small gap is opposite in direction and has a small amplitude. The negative compensation can be appropriately conservative, so Kc - <=1.

[0057] The specific formula for calculating segmented compensation is as follows: In ΔG i When >=0: D i =D0+Kc + ×ΔG i In ΔG i When <0: D i =D0+Kc - ×ΔG i Among them, D i Let D0 be the compensated target displacement of the i-th channel, and Kc be the standard target displacement value. + Kc is the positive compensation coefficient. - The negative compensation coefficient is ΔG. i denoted as the gap deviation value of the i-th channel.

[0058] It needs to be emphasized that Kc + and Kc - The specific value is not directly derived from theoretical formulas, but determined experimentally during the system calibration phase. The calibration goal is to minimize the compensated roundness index Rc (i.e., the range of compression deviations in each channel). The final determined Kc value comprehensively reflects the coupling effect of multiple factors, such as the nonlinear characteristics of the elastic material, the response characteristics of the displacement control system, and the mechanical transmission characteristics of the clamping mechanism, rather than corresponding to a single physical mechanism. In some embodiments, the positive compensation coefficient Kc + The value range is from 1.05 to 1.30, and the negative compensation coefficient Kc - The value ranges from 0.85 to 1.00.

[0059] Continuing with the numerical example above, assuming D0 = 3.00 mm, the calibration determines Kc. + =1.15, Kc - =0.92. Taking channel 1 as an example, ΔG1 = +0.15mm >= 0, then D1 = 3.00 + 1.15 × 0.15 = 3.00 + 0.1725 = 3.1725mm. Taking channel 5 as an example, ΔG5 = -0.12mm < 0, then D5 = 3.00 + 0.92 × (-0.12) = 3.00 - 0.1104 = 2.8896mm. Similarly, the compensated target displacement of all 12 channels can be calculated, forming the compensated target displacement vector D = [D1, D2, ..., D 12 ].

[0060] In some embodiments, to prevent excessive compensation from causing localized stress exceeding limits or rubber damage, the system sets a limit constraint on the compensation amount for each channel. The constraint condition for the compensation amount limit is:

[0061] |ΔD i |=|Kc×ΔG i |<=ΔD max Where, ΔD max The maximum allowable compensation amount is defined by the formulation parameters, with a typical value of 0.3 to 0.5 mm. When the calculated compensation amount for any channel exceeds the limit range, the system determines that the gap deviation of the product has exceeded the compensable range, stops crimping, and issues an alarm.

[0062] Continuing with the numerical examples above, let's assume ΔD max =0.35mm. The compensation amount for channel 1 is |1.15×0.15|=0.1725mm<0.35mm, which meets the limiting requirement. After verification, the compensation amounts for each channel are all within the limiting range, and the limiting verification is passed.

[0063] In some embodiments, after completing the compensation calculation for each channel, the system also performs a consistency check: calculating the range of the target displacement of each channel after compensation, Rd=max(D i )-min(D i If Rd exceeds the set upper limit of the target displacement range after compensation, it indicates that the required compensation difference is too large, the initial state deviation of the product exceeds the effective working range of adaptive compensation, the system determines that it is an uncompensable state and stops the clamping.

[0064] Continuing with the numerical example above, after compensation, the maximum target displacement among the 12 channels is D1 = 3.1725 mm, the minimum target displacement is D5 = 2.8896 mm, and the range Rd = 3.1725 - 2.8896 = 0.2829 mm. Assuming the upper limit of the target displacement range after compensation is 0.50 mm, then Rd = 0.2829 mm < 0.50 mm, and the consistency check passes.

[0065] In this embodiment, the key points of the segmented compensation model are: first, the gap deviation information is quantitatively converted into the displacement compensation amount of each channel, and a feedforward mapping relationship is established between the pre-clamping state and the clamping execution parameters; second, the displacement control system linear tracking deviation caused by the nonlinear characteristics of the elastic material is comprehensively compensated by the positive and negative differential compensation coefficients, and the coefficient values ​​are calibrated to minimize the roundness index; third, the safety is ensured while ensuring the effectiveness of compensation through the dual constraints of amplitude limiting protection and consistency verification.

[0066] Step S104: Send the compensated target displacement vector to the clamping control module, use the independent compensated target displacement of each channel as the displacement control endpoint to perform the clamping action, and collect the actual final displacement value of each channel.

[0067] After successful verification, the system sends the compensated target displacement vector D to the crimping control module. The crimping control module updates the displacement control endpoint of each channel from the previously uniform D0 to an independent D for each channel. i value.

[0068] In some embodiments, each channel is equipped with an independent servo actuator, which can independently control the forward and backward displacement of each channel. The system performs independent displacement closed-loop control for each channel, and each channel reaches its respective target displacement D. i Then stop. This embodiment preferentially adopts this scheme, that is, each channel has independent displacement control capability.

[0069] In other embodiments, for applications with lower production capacity requirements and more relaxed cycle time constraints, each channel shares a single drive source but is differentiated through elastic compensation pads with variable preload. Before crimping, the system adjusts the pre-compression of the elastic pads in each channel by controlling inflation and deflation, thus creating differences in the equivalent stroke endpoints of each channel. The pre-compression of the elastic pads in each channel is determined by the channel's compensation amount ΔD. i =D i -D0 corresponds to the setting (channels with positive compensation increase pre-compression, channels with negative compensation decrease pre-compression), and the mapping relationship between inflation pressure and pre-compression is obtained through pre-calibration; under this scheme, each channel is still equipped with an independent displacement sensor to collect the actual final displacement value D. i,actual .

[0070] This solution is suitable for retrofitting existing crimping equipment that does not have independent channel control capabilities.

[0071] During the clamping process, the system synchronously acquires force and displacement values ​​in real time at high speed (sampling rate not less than 1000Hz) for all channels and records the force-displacement curves for each channel.

[0072] In this embodiment, by performing crimping with independent and differentiated target displacements for each channel, the crimping stroke in the direction of larger gap is appropriately increased and the crimping stroke in the direction of smaller gap is appropriately reduced. Under the condition that the compensation coefficient has been calibrated and optimized, the actual radial compression in each direction tends to be consistent, and the roundness deviation caused by uneven gap is actively eliminated.

[0073] Step S105: Determine the actual radial compression of each channel based on the actual final displacement value and gap deviation vector of each channel, and evaluate the roundness of the product based on the consistency of the actual radial compression of each channel.

[0074] like Figure 5 As shown, after the clamping action is completed, the system reads the actual final displacement value D of each channel. i,actual And perform roundness evaluation.

[0075] The roundness evaluation method of this approach differs fundamentally from traditional methods. In the traditional uncompensated mode, the target displacement of each channel is the same, so the consistency of the actual displacement values ​​of each channel can be directly used as the roundness index. However, in the adaptive compensation mode, the target displacement of each channel is inherently different (Di of each channel is different). i (Since the values ​​are different), the basis for roundness evaluation should not be the consistency of the actual displacement values ​​of each channel, but rather the consistency of the actual radial compression of each channel.

[0076] The formula for calculating the compression deviation value of each channel is as follows: ε i =(D i,actual -ΔG i )-D0 Where, ε i Let D be the compression deviation value of the i-th channel. i,actual Let ΔG be the actual final displacement value of the i-th channel. i Let D be the gap deviation value of the i-th channel, and D0 be the standard target displacement value. The physical meaning of this formula is: D i,actual -ΔG i This represents the effective radial compression displacement of the channel after deducting the gap deviation. Subtracting the standard target compression amount D0 gives the compression deviation of the channel relative to the ideal state.

[0077] Under the condition that the compensation coefficient has been fully calibrated and the displacement control accuracy is good, the εi of each channel should be stably distributed near the reference level determined in the calibration stage, and the difference between channels should not increase with production drift, indicating that the actual radial compression of each channel is consistent.

[0078] The formula for calculating the roundness index Rc after compensation is: Rc=max(ε i )-min(ε i ) Rc represents the range of compression deviation values ​​for each channel, reflecting the consistency of the actual radial compression across all channels. A smaller Rc indicates more uniform compression across all channels and better product roundness.

[0079] If Rc <= the preset roundness tolerance threshold, the product's roundness is determined to be acceptable. If Rc > the roundness tolerance threshold, the product's roundness is determined to be unacceptable.

[0080] Continuing with the numerical example above, let's assume that the actual final displacement of each channel after crimping is D. 1,actual =3.17mm, D 5,actual=2.89mm, etc. Taking channel 1 as an example: ε1=(3.17-0.15)-3.00=0.02mm. Taking channel 5 as an example: ε5=(2.89-(-0.12))-3.00=0.01mm. Assuming the calculation of ε for all 12 channels... i Then, we obtain max(ε) i ) = 0.03 mm, min(ε i If the radius of curvature is -0.02mm, then Rc = 0.03 - (-0.02) = 0.05mm. Assuming the roundness tolerance threshold is 0.08mm, then Rc = 0.05mm < 0.08mm, and the roundness is acceptable.

[0081] In contrast, if crimping is performed directly with a uniform D0=3.00mm without compensation, under the same gap distribution conditions, the ε of each channel... i Adding the tracking deviation to the negative value that is directly equal to the gap deviation, the roundness index Rc will approach the gap deviation range R=0.27mm, far exceeding the tolerance threshold of 0.08mm. This indicates that adaptive compensation reduces the roundness index from approximately 0.27mm to 0.05mm, an improvement of over 80%.

[0082] In some embodiments, the system executes multi-layer decision logic to comprehensively evaluate the crimping quality. The first layer is single-channel positioning determination, which compares the actual final displacement value of each channel with the compensated target displacement of the corresponding channel to determine whether each channel has reached its target position. If |D i,actual -D i If the displacement tolerance (typically 0.02mm) is exceeded, the channel is considered to have an abnormal displacement. The second layer involves single-channel force-displacement curve envelope determination. The measured force-displacement curves of each channel are compared with the qualified envelope, where the qualified envelope is adjusted by corresponding translation based on the compensation amount of each channel. This translation is performed along the displacement axis, with the initial contact segment of the curve adjusted according to the channel gap deviation value ΔG. i Translation, the final segment is compensated by the channel amount Kc×ΔG i Translation. The intermediate section between the starting segment and the ending segment is translated by linear interpolation along the displacement axis; the qualified envelope is generated by statistically expanding the measured force-displacement curves of qualified products in the calibration stage using upper and lower limits.

[0083] The third layer is a roundness evaluation based on the consistency of actual radial compression, calculating the roundness index Rc according to the aforementioned method. The fourth layer is a comprehensive judgment; the system determines the crimping result to be qualified only if all three layers are qualified.

[0084] In this embodiment, roundness evaluation is performed based on the consistency of the actual radial compression of each channel, which correctly reflects the essential quality indicator of sealing uniformity. This adapts to the situation where the target displacement of each channel is different under the adaptive compensation mode. At the same time, the multi-layer judgment logic comprehensively evaluates the crimping quality from three dimensions: displacement accuracy, mechanical properties, and geometric accuracy, ensuring the reliability of the quality judgment.

[0085] Step S106: Based on the compression deviation data after clamping of multiple consecutive products, the compensation coefficient is iteratively corrected.

[0086] like Figure 6 As shown, another core feature of this method is the establishment of a feedback correction loop from the measured data after crimping to the compensation model coefficients, which enables the compensation accuracy to be continuously optimized as the production process progresses.

[0087] The core basis of feedback correction is the balance of compression deviations between the positive compensation channel group and the negative compensation channel group. Specifically, for each qualified product that has completed crimping, the system uses the compression deviation value ε of each channel calculated in step S105. i The following indicators are calculated based on channel grouping:

[0088] Average compression deviation of the positive channel group: ε pos,avg =mean(ε i For all ΔG i Channel i >= 0 Average compression deviation of negative channel group: ε neg,avg =mean(ε i For all ΔG i <0 channel i) Compression deviation channel group balance index: Δε=ε pos,avg -ε neg,avg Here, Δε reflects the systematic difference in effective compression between the positive compensation channel group and the negative compensation channel group. When the compensation coefficients are in optimal calibration, ε for each channel... i The average ε of the positive and negative groups is stably distributed near the calibration reference level, and the difference between the average ε of the positive and negative groups remains at the reference value Δε0 recorded during the calibration phase.

[0089] It should be noted that Δε only controls the balance between positive and negative channel groups, and does not directly control the dispersion between channels within each group (the dispersion within a group is determined by the accuracy of the Kc value determined in the initial calibration stage and the random fluctuations of the product itself). Therefore, Δε = Δε0 is a necessary condition for maintaining the global roundness index Rc at its minimum value, and its sufficiency depends on the following typical operating conditions: the initial calibration has minimized the dispersion within the group, and the characteristic drift in subsequent production is mainly manifested as a common mode shift between positive and negative channel groups (such as the overall change in rubber batch hardness causing all positive channels or all negative channels to be affected in the same direction). Under this premise, maintaining Δε = Δε0 is equivalent to maintaining the optimal Rc state established in the initial calibration. In actual production, typical drift factors such as batch fluctuations in rubber hardness and wear of the crimping mechanism are of the common mode nature between groups, satisfying the above premise.

[0090] When the system exhibits slow drift, Δε will deviate from the reference value Δε0. Specifically: (Δε−Δε0)>0 indicates that the compression of the positive channel group is relatively high compared to the calibration reference system; (Δε−Δε0)<0 indicates that the compression of the negative channel group is relatively high compared to the calibration reference system.

[0091] The system accumulates and statistically analyzes the Δε values ​​of M consecutive products (M is a configurable parameter, typically ranging from 20 to 50 products), and calculates the average compression balance index Δε for the M products. M And calculate its balance deviation (Δε) relative to the calibration reference value. M −Δε0); in |Δε M If −Δε0| exceeds the preset balance offset threshold, a systematic drift trend is determined, triggering the compensation coefficient correction process.

[0092] The correction logic employs an incremental adjustment method, simultaneously making symmetrical corrections to both positive and negative compensation coefficients to reduce the balance deviation (Δε). M −Δε0) drives the coefficient back to zero. The iterative correction formula for the compensation coefficient is:

[0093] Kc + new =Kc + current -α×(Δε M −Δε0) / C pos,avg Kc - new =Kc - current -α×(Δε M −Δε0) / |C neg,avg | Among them, Kc + new Kc is the corrected positive compensation coefficient.+ current Δε is the current positive compensation coefficient, α is the learning rate coefficient, and Δε is the learning rate coefficient. M C represents the average compression balance index for a continuous series of M products. pos,avg The average compensation amount of the positive compensation channel in a continuous series of M products (i.e., Kc of each positive channel) + ×ΔG i (mean), |C neg,avg | represents the absolute value of the average compensation amount for the negative compensation channel. Kc - new and Kc - current The meaning can be deduced by analogy.

[0094] The physical meaning and stability analysis of the corrected formula are as follows: when (Δε M When −Δε0)>0 (the positive channel group is overcompressed relative to the calibration reference), the correction increment is negative, Kc + Reducing the compression of the forward channel decreases, while Kc - Decreasing the compression of the negative channel increases the compression amount, and both factors together drive Δε back to zero. When (Δε) M When −Δε0)<0 (the negative channel group is overcompressed relative to the calibration reference), the correction increment is positive, Kc. + Increasing the value of Kc increases the compression of the forward channel. - Increasing this reduces the compression of the negative channel, similarly driving the balance deviation back to zero. Since the correction direction is always related to (Δε) M -Δε0) has the opposite sign (negative feedback). Within the range of α (0.1 to 0.3), the system stably converges to the equilibrium point of Δε = Δε0. That is, under the typical operating condition of the aforementioned inter-group common mode drift, this equilibrium point corresponds to maintaining the optimal Rc state established by the initial calibration.

[0095] If there is no negative (or positive) channel in the gap deviation vector of a product, or if the average compensation amount of any channel group is less than the preset minimum compensation threshold, the product will not be included in the balance statistics; if the number of valid samples in the statistical window is less than M, the statistics will be extended without triggering coefficient correction.

[0096] In some embodiments, the learning rate coefficient α typically ranges from 0.1 to 0.3 to control the correction rate and avoid over-adjustment that could cause system oscillations. The number of consecutive products M typically ranges from 20 to 50 to ensure statistical reliability.

[0097] The corrected coefficients need to meet a preset reasonable range constraint. In some embodiments, the reasonable range constraint for the positive compensation coefficients is Kc. + For the negative compensation coefficient ∈ [1.0, 1.5], the reasonable range constraint is Kc.- ∈[0.7,1.0]. If the calculated result exceeds a reasonable range, the system will not perform the correction and will generate an alarm, indicating that there may be a systemic change in mechanical state requiring manual intervention for investigation. Each coefficient correction records the value before correction, the value after correction, the statistical basis data that triggered the correction, and the correction timestamp, for process personnel to trace and analyze.

[0098] For example, suppose the current Kc + =1.15, Kc - =0.92, the reference value Δε0 recorded during the calibration phase of this product model is 0.000mm (the specific value of Δε0 depends on the statistical results of the calibration measurements; here, a zero value is used for illustrative purposes). After statistical analysis of 30 consecutive products, the average compression deviation ε of the forward channel group is... pos,avg =0.005mm, average compression deviation ε of negative channel group neg,avg =0.018mm, then the compression balance index Δε M =0.005-0.018=-0.013mm. Therefore, the balance deviation (Δε) M −Δε0)=−0.013mm, since |Δε M −Δε0|=0.013mm>balance offset threshold 0.010mm, triggering correction. This negative value indicates that the negative channel group is relatively overcompressed, i.e., there is an imbalance between the positive and negative channels.

[0099] Assume C pos,avg =0.10mm, |C neg,avg Given |=0.08mm and learning rate α=0.2, the corrected calculation is as follows:

[0100] Kc + new =1.15 - 0.2 × (-0.013) / 0.10 = 1.15 + 0.026 = 1.176 Kc - new =0.92 - 0.2 × (-0.013) / 0.08 = 0.92 + 0.0325 = 0.9525 Due to Kc + new =1.176 is in [1.0, 1.5], Kc - new =0.9525 is within the range of [0.7, 1.0], both within a reasonable range, and the system performs a correction. After the correction, the compression amount of the forward channel increases (due to Kc). + Increasing the positive channel leads to an increase in target displacement), while decreasing the negative channel compression (due to Kc). -Increasing the negative channel compensation amount leads to an increase in absolute value and a decrease in effective compression displacement. It is expected that Δε will return to the reference value Δε0, that is, the compression deviation balance between the two channels will be restored to the calibration reference state, and the roundness index Rc will decrease.

[0101] like Figure 7 As shown, after adaptive compensation control, the distribution of the product roundness index Rc decreased from the mean of 0.28 mm to 0.04 mm and the standard deviation decreased from 0.08 mm to 0.015 mm compared with the control group without compensation. The roundness pass rate (Rc<=0.08 mm) increased from about 72% to about 98%, indicating that the compensation method effectively improved the consistency of product roundness.

[0102] In this embodiment, by using the inter-group balance index Δε of the compression deviation channel as a feedback quantity for iterative correction of the compensation coefficient, a clear causal chain of feedback quantity (Δε) → correction direction → roundness maintenance (Rc does not deteriorate) is established. Under the typical operating condition where the initial calibration has optimized the intra-group dispersion and the production drift is an inter-group common mode offset, Δε = Δε0 corresponds to the optimal Rc state. When Δε deviates from the reference value, the correction direction will balance the deviation quantity (Δε). M The system converges stably under the given α value by driving the return to zero (negative feedback) with −Δε0. This mechanism enables the compensation model to have self-learning ability to resist slow common-mode drift, and can automatically adapt to factors such as batch fluctuations in airbag rubber hardness and gradual changes in characteristics caused by wear of the buckling mechanism, maintaining the compensation accuracy established in the initial calibration over a long period of time.

[0103] In some embodiments, the system performs initial calibration of the compensation model when first launching a new product model or restarting after a long period of downtime. The calibration process includes three stages:

[0104] The first stage is the establishment of an uncompensated baseline: the system performs standard crimping on no fewer than 30 products with a standard target displacement D0, and simultaneously records the pre-crimping clearance deviation ΔG for each product. i And the actual displacement value D of each channel after crimping i,actual Calculate the compression deviation ε for each channel. i =(D i,actual -ΔG i -D0, statistically analyze the average ε difference between the positive and negative channel groups. During the calibration phase, roundness evaluation uses an offline roundness measurement device (such as a roundness meter or coordinate measuring machine) to independently measure the outer contour of the finished product to obtain the measured roundness index; if the positive channel system systematically shows undercompression according to the measured roundness evaluation, then Kc is required. + >1. Compensation is performed. By gradually testing different Kc values ​​and statistically analyzing the corresponding measured roundness distributions, the Kc value that minimizes the measured roundness is determined. + and Kc -Initial value, and record the value at ε when calibration is complete. i The baseline value Δε0 of the channel group balance index obtained by calculation and statistics is used as a reference benchmark for online feedback.

[0105] The second stage is compensation verification and adjustment: The system starts the adaptive compensation mode with the initial Kc value and performs compensation pressure on several products. The roundness index Rc distribution after compensation is compared with the roundness distribution in the uncompensated stage. If the mean and standard deviation of the Rc distribution are significantly better than those in the uncompensated stage, the compensation is confirmed to be effective. If there is a systematic balance shift (the Δε statistical value continues to drift), the coefficient is corrected according to the aforementioned incremental correction method, and the verification is repeated until the Δε statistical value is stable and the Rc distribution meets the process requirements. The finally confirmed Δε statistical value is then updated to the baseline value Δε0.

[0106] The third stage is the confirmation of the formal production model: after calibration, the final Kc will be determined. + Kc - The values ​​and related parameters are written into the product formula database, and the system enters formal production mode.

[0107] The following describes the multi-channel adaptive compensation control device provided in the embodiments of the present invention.

[0108] like Figure 8 As shown, this embodiment of the invention provides a multi-channel adaptive compensation control device for the air spring clamping process, including the following functional modules: The gap detection module acquires gap measurements at multiple circumferential channels between the clamping ring and the main body before clamping. Based on these measurements and a preset standard gap value, it determines the gap deviation value for each channel, forming a gap deviation vector. The gap detection module connects to the vision inspection system via an image data interface, receiving high-resolution image data output by the vision inspection system and executing image processing algorithms to extract the gap values ​​for each channel. The gap detection module also includes a validity verification submodule, which performs range and maximum offset verification on the gap deviation vector, outputting a shutdown alarm signal if the verification fails.

[0109] The compensation calculation module is used to determine the target displacement after compensation for each channel based on the gap deviation vector and using the gap deviation-displacement compensation mapping model, thus forming the target displacement vector after compensation. The compensation calculation module receives gap deviation vector data from the gap detection module, reads compensation coefficients and related parameters from the recipe database, and performs segmented compensation calculations. The compensation calculation module also includes a limiting processing submodule and a consistency verification submodule, used for safety limiting of the compensation amount and range verification of the target displacement vector after compensation, respectively.

[0110] The clamping execution control module is used to execute the clamping action with the independent compensated target displacement of each channel as the displacement control endpoint, and to collect the actual final displacement value of each channel. The clamping execution control module receives the compensated target displacement vector from the compensation calculation module, sends independent displacement command signals to the servo drivers of each channel, and simultaneously collects force and displacement data in real time through the displacement and force sensors of each channel.

[0111] The roundness evaluation module determines the actual radial compression of each channel based on its final displacement value and gap deviation vector, and evaluates the product roundness based on the consistency of the actual radial compression. The roundness evaluation module receives the final displacement data of each channel from the crimping execution control module, obtains the gap deviation vector from the gap detection module, and calculates the compression deviation value ε for each channel. i The roundness index Rc after compensation is calculated, multi-level judgment logic is executed, and the comprehensive quality judgment result is output.

[0112] The model correction module performs statistical analysis of the channel-to-channel balance of compression deviation values ​​for multiple consecutive qualified products. When a systematic balance shift is detected, it triggers a compensation coefficient correction process and updates the corrected coefficients to the compensation calculation module. The model correction module is connected to the roundness evaluation module via a data interface and continuously receives the channel compression deviation values ​​ε for each product. i According to the positive channel group (ΔG) i >=0) and negative channel group (ΔG) i <0) Calculate the average compression deviation and determine the compression balance index Δε. When the balance deviation of M consecutive products is |Δε M When −Δε0| exceeds the balance offset threshold, the model correction module performs incremental correction calculations and writes the results to the coefficient storage area of ​​the compensation calculation module.

[0113] The electronic device provided in the embodiments of the present invention is described below.

[0114] like Figure 9 As shown, an embodiment of the present invention provides an electronic device, including a memory and a processor coupled to the memory. The memory stores computer program instructions, and the processor is configured to execute the aforementioned multi-channel adaptive compensation control method based on the instructions in the memory.

[0115] In some embodiments, the electronic device is an industrial control computer or a programmable logic controller (PLC), and the processor communicates with the vision inspection system and the servo drives of each channel via an industrial Ethernet or fieldbus. The memory stores a product formula database, including the standard gap value G0, standard target displacement value D0, and compensation coefficient Kc corresponding to each product model. + and Kc -Data includes the baseline value Δε0 of the inter-channel balance index, various threshold parameters, historical statistical data on compression deviation balance, and coefficient correction records.

[0116] The electronic device also includes a communication interface for data interaction with external human-machine interaction devices, data management systems, and host computer management systems, enabling functions such as parameter configuration, status monitoring, data recording, and quality traceability.

[0117] The air spring clamping system provided in the embodiments of the present invention is described below.

[0118] like Figure 10 As shown, an embodiment of the present invention provides an air spring clamping system, including a vision inspection system, a multi-lobed clamping mechanism, and a controller.

[0119] The vision inspection system includes a high-resolution industrial camera and lighting system, mounted above the crimping station, to acquire high-resolution images of the circumferential direction between the crimping ring and the main body. The vision inspection system transmits the acquired image data to the controller via an image data interface.

[0120] The multi-lobed clamping mechanism has multiple clamping channels evenly distributed along the circumference, each channel equipped with an independent displacement sensor and force sensor. In some embodiments, each channel is also equipped with an independent servo actuator, capable of independently controlling the radial displacement of each channel. The signals from each sensor are connected to the acquisition port of the controller via signal lines, and each servo actuator is connected to the control output port of the controller via drive signal lines.

[0121] The controller is coupled to the vision inspection system and the multi-lobed clamping mechanism and is configured to execute the aforementioned multi-channel adaptive compensation control method. The controller receives image data from the vision inspection system, performs gap detection and compensation calculation, sends differentiated target displacement commands to the servo actuators of each channel, and collects the force and displacement feedback signals of each channel in real time for closed-loop control and quality judgment.

[0122] In some embodiments, the crimping system is applied in the production line of commercial vehicle seat air spring assemblies. After the airbag assembly and crimping ring placement are completed at the upstream station, the product is conveyed to the crimping station and fixed by positioning fixtures. The controller automatically initiates the vision inspection process, completing a fully automated control process of gap measurement, compensation calculation, and crimping execution. After crimping is completed, the quality judgment result is output and the product data is recorded to the data management system.

[0123] It should be noted that the execution order of the method steps described in the embodiments of the present invention is not limited to the order described above. In the absence of logical contradictions, the order of each step can be adjusted appropriately according to actual needs. For example, although validity verification and amplitude limiting are described as independent steps in the above embodiments, they can be executed as internal sub-processes within the compensation calculation step in actual implementation.

[0124] Those skilled in the art will understand that all or part of the steps in the above embodiments can be performed by a computer program, which can be stored in a computer-readable storage medium. Computer-readable storage media include, but are not limited to, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0125] Those skilled in the art will also understand that the technical features provided in the above embodiments can be combined arbitrarily, and as long as there is no logical contradiction between the technical features, the resulting combination schemes all fall within the protection scope of the present invention.

[0126] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the appended claims.

Claims

1. An air spring crimping process compensation control method, characterized by, include: Before clamping, obtain the gap measurement values ​​at multiple channel positions in the circumferential direction between the clamping ring and the main body. Based on the gap measurement values ​​and the preset standard gap value, determine the gap deviation value of each channel and form a gap deviation vector. Based on the gap deviation vector, the compensation target displacement of each channel is determined using the gap deviation-displacement compensation mapping model, forming a compensation target displacement vector; wherein, the gap deviation-displacement compensation mapping model includes: the mapping relationship between the gap deviation value of each channel and the preset compensation coefficient and standard target displacement value, and the compensation target displacement is a differentiated target value obtained by superimposing the compensation coefficient and the corresponding channel gap deviation value on the standard target displacement value; The compensated target displacement vector is sent to the clamping control module, and the clamping action is executed with the independent compensated target displacement of each channel as the displacement control endpoint, and the actual final displacement value of each channel is collected. Based on the actual final displacement value of each channel and the gap deviation vector, the actual radial compression of each channel is determined, and the roundness of the product is evaluated based on the consistency of the actual radial compression of each channel. The evaluation of product roundness based on the consistency of actual radial compression of each channel includes: calculating the compression deviation value of each channel according to the actual final displacement value of each channel, the gap deviation value of the corresponding channel, and the standard target displacement value; calculating the range of compression deviation values ​​of all channels as the roundness index after compensation; and determining that the product roundness is qualified if the roundness index after compensation does not exceed the preset roundness tolerance threshold.

2. The method as described in claim 1, characterized in that, The gap deviation-displacement compensation mapping model is a piecewise compensation model, and the compensation coefficients include positive compensation coefficients and negative compensation coefficients; The method of using the gap deviation-displacement compensation mapping model to determine the compensated target displacement of each channel includes: If the gap deviation value of the i-th channel is greater than or equal to zero, the compensated target displacement of the i-th channel is determined to be equal to the sum of the product of the standard target displacement value, the positive compensation coefficient, and the gap deviation value of the channel. If the gap deviation value of the i-th channel is less than zero, the compensated target displacement of the i-th channel is determined to be equal to the sum of the product of the standard target displacement value, the negative compensation coefficient, and the gap deviation value of the channel. Wherein, the positive compensation coefficient is greater than 1, the negative compensation coefficient is less than or equal to 1, and the differential setting of the positive compensation coefficient and the negative compensation coefficient is based on the nonlinear stiffness characteristics of the elastic airbag material in different compression sections.

3. The method as described in claim 1, characterized in that, Before determining the compensated target displacement of each channel based on the gap deviation vector, the method further includes: The validity of the gap deviation vector is verified. The validity verification includes: calculating the range of all channel gap deviation values ​​and the maximum offset among the absolute values ​​of each channel gap deviation value; if the range exceeds a preset upper limit threshold or the maximum offset exceeds a preset maximum offset threshold, determining that the gap deviation vector has failed the validity verification, stopping the pressure and triggering an alarm; if the range does not exceed the upper limit threshold and the maximum offset does not exceed the maximum offset threshold, determining that the gap deviation vector has passed the validity verification and proceeding to the compensation calculation stage.

4. The method as described in claim 2, characterized in that, Also includes: The compensation amount for each channel is limited; The limiting process includes: calculating the absolute value of the compensation amount for each channel; if the absolute value of the compensation amount for any channel exceeds the preset maximum allowable compensation amount, determining that the gap deviation of the product exceeds the compensable range, stopping the clamping and triggering an alarm. It also includes: performing a consistency check on the compensated target displacement vector, calculating the range of the target displacement of each channel in the compensated target displacement vector, and if the range exceeds the preset upper limit of the range of the compensated target displacement, determining that the product's initial state deviation exceeds the effective working range of adaptive compensation, and stopping the pressure clamping.

5. The method as described in claim 1, characterized in that, The consistency evaluation of product roundness based on the actual radial compression of each channel also includes executing multi-layer judgment logic: The first layer, single-channel positioning determination: compare the actual final displacement value of each channel with the compensated target displacement of the corresponding channel. If the absolute value of the difference between the two exceeds the preset displacement positioning tolerance, the positioning of that channel is determined to be abnormal. The second layer is the single-channel force-displacement curve envelope determination: the measured force-displacement curve of each channel is compared with the qualified envelope, wherein the qualified envelope is adjusted accordingly based on the compensation amount of each channel. The third layer, roundness evaluation: Calculate the compensated roundness index based on the consistency of the actual radial compression of each channel and compare it with the roundness tolerance threshold; Fourth layer, comprehensive judgment: if the first layer determines that all channels are qualified, the second layer determines that all channels are qualified, and the third layer determines that the roundness is qualified, then the current crimping result is determined to be qualified.

6. The method as described in claim 1, characterized in that, Also includes: For multiple consecutive qualified products that have completed the crimping process, a balance statistical analysis is performed by grouping them by channel based on the compression deviation value of each channel of each product and the gap deviation vector. The channel grouping includes: classifying channels with gap deviation values ​​greater than or equal to zero into a positive compensation channel group, and classifying channels with gap deviation values ​​less than zero into a negative compensation channel group; Calculate the average compression deviation of the positive channel group and the average compression deviation of the negative channel group for each product, and determine the compression deviation balance index between the channel groups as the difference between the average compression deviation of the positive channel group and the average compression deviation of the negative channel group. The compression deviation channel group balance index of M consecutive products is accumulated and statistically analyzed to calculate the average compression balance index of the M products and its balance deviation relative to the preset calibration reference value; if the absolute value of the balance deviation exceeds the preset balance offset threshold, the compensation coefficient correction process is triggered. The compensation coefficient is iteratively corrected using an incremental adjustment method, which includes subtracting the product of the learning rate coefficient and the ratio of the balance deviation to the average compensation of the corresponding channel group from the current compensation coefficient, and obtaining the corrected compensation coefficient. If the corrected compensation coefficient meets the preset reasonable range constraint, it is determined to perform the correction and update the compensation coefficient; if the corrected compensation coefficient exceeds the reasonable range constraint, it is determined not to perform the correction and generate an alarm.

7. A multi-channel adaptive compensation control device for the air spring clamping process, characterized in that, The method described in any one of claims 1-6 includes: The gap detection module is used to obtain the gap measurement values ​​at multiple channel positions in the circumferential direction between the clamping ring and the main body before clamping. Based on the gap measurement values ​​and the preset standard gap values, the gap deviation value of each channel is determined to form a gap deviation vector. The compensation calculation module is used to determine the target displacement after compensation for each channel based on the gap deviation vector and using the gap deviation-displacement compensation mapping model, thereby forming a target displacement vector after compensation. The gap deviation-displacement compensation mapping model includes a mapping relationship between the gap deviation value of each channel and a preset compensation coefficient and a standard target displacement value. The target displacement after compensation is a differentiated target value obtained by superimposing the compensation coefficient and the corresponding channel gap deviation value on the standard target displacement value. The clamping execution control module is used to execute the clamping action with the independent compensated target displacement of each channel as the displacement control endpoint, and to collect the actual final displacement value of each channel. The roundness evaluation module is used to determine the actual radial compression of each channel based on the actual final displacement value of each channel and the gap deviation vector, calculate the range of compression deviation values ​​of all channels as the roundness index after compensation, and evaluate the roundness of the product based on the consistency of the actual radial compression of each channel. The model correction module is used to perform statistical analysis on the compression deviation balance of multiple consecutive qualified products after crimping, grouped by channel. Channels with gap deviation values ​​greater than or equal to zero are classified as positive compensation channel groups, and channels with gap deviation values ​​less than zero are classified as negative compensation channel groups. The average compression deviation of the positive channel group and the average compression deviation of the negative channel group for each product are calculated to determine the balance index between compression deviation channel groups. The balance index between the compression deviation channel groups of multiple consecutive products is accumulated and statistically analyzed to obtain the average compression balance index. The balance deviation of the average compression balance index relative to the preset calibration benchmark value is determined. If the absolute value of the balance deviation exceeds the preset balance offset threshold, the compensation coefficient is iteratively corrected using an incremental adjustment method, and the corrected compensation coefficient is updated to the compensation calculation module.

8. An air spring clamping system, characterized in that, include: A vision inspection system is used to acquire high-resolution images of the circumferential direction between the clamping ring and the main body component; A multi-lobed clamping mechanism has multiple clamping channels evenly distributed along the circumference. Each channel is equipped with an independent displacement sensor and a force sensor. The displacement sensor is used to collect the radial displacement value of the corresponding channel in real time, and the force sensor is used to collect the clamping force value of the corresponding channel in real time. The displacement value and the force value are used to construct the force-displacement curve of each channel. A controller, coupled to the vision inspection system and the multi-lobed clamping mechanism, is configured to perform the method as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, include: Memory; And a processor coupled to the memory, the processor being configured to perform the method as described in any one of claims 1 to 6 based on instructions stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images