Intelligent assembly system and method for self-adaptive variable-diameter production of shuttle peg sleeve
By identifying burr features through image acquisition and combining machine learning to analyze the impact and damage of clamping, clamping parameters are optimized, solving the problem of poor assembly effect in the production of bobbin sleeves with variable diameter, and realizing efficient assembly in adaptive variable diameter production.
Patent Information
- Application Number
- CN202511119497.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-11
- Publication Date
- 2025-11-18
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In existing technologies, the assembly effect is poor and the yield is low when producing bobbin sleeves with variable diameters. Traditional systems lack dynamic sensing capabilities, which causes the clamping parameters to be unsuitable for changes in the inner diameter, affecting production quality and stability.
An image acquisition and recognition module is used to dynamically capture the characteristics of variable diameter burrs. A clamping influence and damage analysis model is constructed by combining machine learning. Clamping parameters are optimized through clamping and polishing analysis to achieve adaptive variable diameter production.
It significantly improves the yield and dimensional consistency of variable diameter bobbin assembly, enhances assembly stability and polishing precision, and adapts to adaptive assembly of various inner diameter specifications.
Smart Images
Figure CN120962452A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of data processing, in particular to an intelligent assembly system and method for adaptive variable-diameter production of bobbin cases. BACKGROUND
[0002] In the field of sewing equipment manufacturing, the bobbin case as a core transmission component directly affects the working stability and service life of the sewing machine. With the growth of personalized customization demand, variable-diameter bobbin case production has become an industry trend, that is, continuously producing bobbin cases with different inner diameters on the same production line. The existing technology mainly relies on a fixed parameterized assembly system to process bobbin cases by presetting clamping force. However, variable-diameter production leads to nonlinear characteristics of clamping demand, and the traditional system still uses uniform clamping parameters due to the lack of dynamic sensing capability, resulting in poor production assembly effect and reduced yield. SUMMARY
[0003] The present application provides an intelligent assembly system and method for adaptive variable-diameter production of bobbin cases to solve the technical problem of poor assembly effect in the variable-diameter production of bobbin cases in the prior art.
[0004] In view of the above problems, the present application provides an intelligent assembly system and method for adaptive variable-diameter production of bobbin cases.
[0005] In a first aspect, the present application provides an intelligent assembly system for adaptive variable-diameter production of bobbin cases, comprising: An image acquisition and recognition module is configured to acquire images of bobbin cases produced according to varying inner diameters and recognize variable-diameter burr characteristic parameters during variable-diameter production of bobbin cases.
[0006] A clamping damage influence analysis module is configured to randomly generate clamping parameters for clamping and polishing of bobbin cases with varying inner diameters, combine the variable-diameter burr characteristic parameters, and perform clamping influence analysis and clamping damage analysis to obtain clamping influence parameters and clamping damage parameters.
[0007] A clamping and polishing analysis module is configured to perform clamping and polishing analysis according to the clamping parameters and variable-diameter burr characteristic parameters to obtain clamping and polishing parameters.
[0008] A clamping parameter optimization module is configured to calculate an assembly score based on the clamping influence parameters, clamping damage parameters, and clamping and polishing parameters, iteratively optimize the clamping parameters, obtain optimal clamping parameters, and perform assembly processing of variable-diameter bobbin cases.
[0009] In a second aspect, the present application provides an intelligent assembly method for adaptive variable-diameter production of bobbin cases, comprising: During variable-diameter production of bobbin cases, images of bobbin cases produced according to varying inner diameters are acquired, and variable-diameter burr characteristic parameters are recognized.
[0010] The clamping parameters of the clamping and polishing of the variable diameter core sleeve are randomly generated, the clamping influence analysis and clamping damage analysis are performed in combination with the variable diameter burr characteristic parameters, and clamping influence parameters and clamping damage parameters are obtained.
[0011] According to the clamping parameters and the variable diameter burr characteristic parameters, clamping polishing analysis is performed to obtain clamping polishing parameters.
[0012] Based on the clamping influence parameters, the clamping damage parameters and the clamping polishing parameters, an assembly score is calculated, the clamping parameters are iteratively optimized to obtain optimal clamping parameters, and assembly processing of the variable diameter core sleeve is performed.
[0013] One or more technical solutions provided in the present application have at least the following technical effects or advantages: The present application provides an intelligent assembly system and method for adaptive variable diameter production of a core sleeve. By dynamically capturing variable diameter burr characteristics and constructing a collaborative optimization mechanism of clamping parameters-damage-polishing, the yield and size consistency of the variable diameter core sleeve assembly are significantly improved. Compared with the traditional method, the technical solution provided by the present application overcomes the adaptability defects of the fixed parameter assembly system in dealing with variable diameter production, and achieves the technical effect of adaptively assembling core sleeves of multiple inner diameter specifications without manual intervention, while simultaneously improving the assembly stability, safety and polishing precision. BRIEF DESCRIPTION OF DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.
[0015] Figure 1 A structural schematic diagram of an intelligent assembly system for adaptive variable diameter production of a core sleeve is provided for the embodiments of the present application.
[0016] Figure 2 A flowchart of an intelligent assembly method for adaptive variable diameter production of a core sleeve is provided for the embodiments of the present application.
[0017] In the drawings, the components represented by the numbers are described as follows: Image acquisition and recognition module 100, clamping damage influence analysis module 200, clamping polishing analysis module 300, clamping parameter optimization module 400. DETAILED DESCRIPTION
[0018] The application provides an intelligent assembly system and method for self-adaptive diameter change production of a bobbin cover, and aims to solve the technical problem of poor assembly effect in the diameter change production of the bobbin cover in the prior art.
[0019] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.
[0020] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server comprising a series of steps or units does not have to be limited to the clearly listed steps or units, but can include other steps or modules that are not clearly listed or inherent to the process, method, product or device.
[0021] Embodiment one, as shown in the application provides an intelligent assembly system for self-adaptive diameter change production of a bobbin cover, wherein the system comprises: Figure 1 An image acquisition and identification module 100 is configured to acquire images of the bobbin cover produced according to the changing inner diameter during the diameter change production of the bobbin cover, and identify and obtain the diameter change burr feature parameters, including: During the diameter change production of the bobbin cover, the image acquisition and identification module 100 acquires images of the bobbin cover produced according to the changing inner diameter, and obtains the diameter change image. During the diameter change production of the bobbin cover, the image acquisition and identification module 100 acquires images of the bobbin cover produced according to the changing inner diameter, and obtains the diameter change image.
[0022] The diameter change image is input into the burr feature identifier, and the diameter change burr feature parameters are identified and output.
[0023] The training step of the burr feature identifier comprises: Based on the bobbin cover production data in the historical time, a sample image set is acquired, and the burr feature parameters of the bobbin cover in different sample images are acquired, to obtain a sample burr feature parameter set, wherein each sample burr feature parameter comprises a burr size.
[0024] The burr feature identifier is constructed based on deep learning.
[0025] The sample image set and the sample burr feature parameter set are used to iteratively train and optimize the burr feature identifier, and the training is completed after the test loss meets the convergence requirement.
[0026] During the diameter change production of the bobbin cover, the inner diameter fluctuation causes the burr shape to be distributed in a nonlinear and random manner, and the traditional method adopts manual visual inspection, which cannot accurately quantify the dynamic burr feature.
[0027] In the embodiment of the application, during the production of the bobbin sleeve with variable diameter, a camera is used to collect images of the bobbin sleeve produced according to the variable inner diameter.
[0028] Based on the production data of the bobbin sleeve in the historical time, a sample image set is collected. For example, images produced at different diameters in the past 30 days are collected, and a sample image set is obtained by integration. The burr feature parameters of the bobbin sleeve in different sample images are collected to obtain a sample burr feature parameter set. Each sample burr feature parameter includes a burr size. Optionally, the length of the burr is used to represent the burr size, and the unit is mm.
[0029] Based on deep learning, a burr feature recognizer is constructed. For example, a 4-layer structure is used, wherein the input layer is used to receive the image of the bobbin sleeve, the first hidden layer uses 64 3×3 convolution kernels, the second hidden layer uses 32 3×3 convolution kernels, the output layer outputs the recognized burr size, and the loss function uses the mean square error function.
[0030] The sample image set and the sample burr feature parameter set are used to iteratively train and optimize the burr feature recognizer. After the test loss meets the convergence requirement, for example, the error of the burr size output by the input sample image is within 0.1 mm, the training of the burr feature recognizer is completed.
[0031] The variable diameter image is input into the burr feature recognizer to obtain the variable diameter burr feature parameter.
[0032] Through the training of the deep learning burr feature recognizer, the background noise and the real burr information can be effectively separated. The variable diameter burr feature parameter generated thereby provides an input benchmark with clear physical meaning for subsequent clamping optimization and lays the foundation for adaptive control perception.
[0033] The clamping damage influence analysis module 200 is used to randomly generate clamping parameters for the clamping and polishing of the bobbin sleeve with variable inner diameter, combine the variable diameter burr feature parameter, and perform clamping influence analysis and clamping damage analysis to obtain clamping influence parameters and clamping damage parameters, including: The clamping parameters for the clamping and polishing of the bobbin sleeve with variable inner diameter are randomly generated, wherein the clamping parameters include clamping force parameters.
[0034] The clamping parameters and the variable diameter burr feature parameters are input into the clamping influence analysis branch and the clamping damage analysis branch to obtain clamping influence parameters and clamping damage parameters, wherein the clamping influence parameters include clamping influence size, and the clamping damage parameters include damage size.
[0035] The clamping influence analysis branch and the clamping damage analysis branch are trained by the following steps: According to historical production data of the inner diameter bobbin sleeve, a sample clamping parameter set, a sample variable diameter burr feature parameter set are collected, and a size precision change value under clamping of different sample clamping parameters and sample variable diameter burr feature parameters is collected to obtain a sample clamping influence parameter set.
[0036] The damage size of the bobbin sleeve after clamping under different sample clamping parameters and sample variable diameter burr feature parameters is collected to obtain a sample clamping damage parameter set.
[0037] A clamping influence analysis branch and a clamping damage analysis branch based on machine learning are constructed.
[0038] The sample clamping parameter set and the sample variable diameter burr feature parameter set are used as input data, the sample clamping influence parameter set and the sample clamping damage parameter set are used as output data, and the clamping influence analysis branch and the clamping damage analysis branch are iteratively trained and optimized, and the training is completed after the test loss meets the convergence requirement.
[0039] The traditional assembly system uses single target parameter optimization such as only minimizing clamping deformation, ignoring the coupling effect of burr and clamping parameters: on the one hand, fixed clamping force induces stress concentration in the thin-walled area, which may exacerbate crack propagation at the burr root; on the other hand, the burr itself changes the local contact area, causing mismatch of clamping force distribution.
[0040] In the embodiment of the application, clamping parameters for clamping and polishing of the variable inner diameter bobbin sleeve are randomly generated, wherein the clamping parameters include various clamping parameters, such as the clamping method, which can be hydraulic clamping or mechanical clamping; the size of the clamping force, which is in N, such as 5N. Different clamping parameters will affect the clamping effect, and then affect the assembly effect.
[0041] According to historical production data of the inner diameter bobbin sleeve, a sample clamping parameter set, a sample variable diameter burr feature parameter set are collected, and a size precision change value under clamping of different sample clamping parameters and sample variable diameter burr feature parameters is collected to obtain a sample clamping influence parameter set.
[0042] After clamping under different sample clamping parameters and sample variable diameter burr feature parameters, the damage size of the bobbin sleeve is collected, and optionally the amplitude of the change in the diameter of the bobbin sleeve after clamping is used as the damage size, for example, the diameter of the bobbin sleeve is reduced by 0.2mm after clamping, and 0.2mm is used as the bobbin sleeve damage parameter. A sample clamping damage parameter set is integrated.
[0043] Based on machine learning, a clamping influence analysis branch and a clamping damage analysis branch are constructed, wherein the clamping influence analysis branch adopts a 3-layer structure, the input layer receives clamping parameters and variable-diameter burr characteristics, the hidden layer adopts 32 nodes and uses a ReLU function for activation, and the output layer outputs clamping influence parameters. The clamping damage analysis branch adopts a 3-layer structure, the input layer receives clamping parameters and variable-diameter burr characteristics, the hidden layer adopts 32 nodes and uses a ReLU function for activation, and the output layer outputs clamping damage parameters.
[0044] The sample clamping parameter set and the sample variable-diameter burr characteristic parameter set are used as input data, and the sample clamping influence parameter set and the sample clamping damage parameter set are used as output data, respectively. The clamping influence analysis branch and the clamping damage analysis branch are respectively subjected to supervised iterative training and optimization. The training is completed when the test loss meets the convergence requirement, and the trained clamping influence analysis branch and clamping damage analysis branch are obtained.
[0045] The clamping parameters and the variable-diameter burr characteristic parameters are input into the clamping influence analysis branch and the clamping damage analysis branch to obtain clamping influence parameters and clamping damage parameters, wherein the clamping influence parameters include clamping influence sizes, and the clamping damage parameters include damage sizes.
[0046] The application constructs a clamping influence and damage double-branch analysis model. Through machine learning, a coupling mapping of burr characteristic parameters, clamping parameters, clamping influence and clamping damage is established. The model can learn the stress influence law from historical data and accurately output clamping influence sizes and damage sizes, thereby providing accurate optimization data reference for adaptive assembly.
[0047] The clamping polishing analysis module 300 is used for clamping polishing analysis according to the clamping parameters and the variable-diameter burr characteristic parameters to obtain clamping polishing parameters, including: The clamping polishing classifier is obtained, wherein the clamping polishing classifier is constructed by machine learning.
[0048] The clamping parameters and the variable-diameter burr characteristic parameters are input into the clamping polishing classifier to obtain clamping polishing parameters through classification output, wherein the clamping polishing parameters include polishing rates.
[0049] The polishing efficiency of the variable-diameter shuttle core sleeve also depends on the matching degree of the burr distribution and the clamping state. However, the traditional process usually processes the two in a fragmented manner. The polishing effect is often considered based on ideal clamping working conditions and cannot consider the influence caused by clamping deviation due to actual clamping force fluctuation. For example, when the burr is located in the variable-diameter transition zone, rigid clamping causes the workpiece to tilt slightly, causing the polishing tool to be misaligned with the shuttle core sleeve or to slip, which may cause local polishing deficiency.
[0050] In the embodiment of the application, a clamping polishing classifier is constructed by machine learning.
[0051] Exemplarily, a 3-layer structure is adopted to construct the clamping polishing classifier, wherein the input layer is used to receive the clamping parameters and the variable-diameter burr feature parameters, the hidden layer adopts 32 nodes and uses the ReLU function for activation, and the output layer is used to output the clamping polishing parameters obtained by classification, and the clamping polishing parameters include a polishing rate. The polishing rate is a parameter for representing the polishing effect, for example, only 80% of the surface of the core sleeve is polished, and the polishing rate is 80%.
[0052] The sample clamping parameter set and the sample variable-diameter burr feature parameter set are labeled, and the labeling content is the corresponding clamping polishing parameter. The sample clamping polishing parameter set is obtained by integrating the labeling content.
[0053] The constructed clamping polishing classifier is supervised trained until convergence by using the sample clamping parameter set, the sample variable-diameter burr feature parameter set and the sample clamping polishing parameter set, for example, the error of the output clamping polishing parameter is within ±3%, that is, the clamping polishing classifier is trained.
[0054] The clamping parameters and the variable-diameter burr feature parameters are input into the clamping polishing classifier, and the clamping polishing parameters are obtained by classification output.
[0055] By constructing the clamping polishing classifier, the burr feature and the clamping parameter are fused, and the polishing rate under the actual working condition is predicted. The clamping polishing classifier learns the complex nonlinear relationship between the clamping state and the polishing effect in the historical data, outputs the polishing parameter of the current clamping condition, quantitatively represents the influence of the clamping effect on the polishing, and provides a basis for subsequent optimization of the clamping parameter.
[0056] The clamping parameter optimization module 400 is used to calculate an assembly score based on the clamping influence parameter, the clamping damage parameter and the clamping polishing parameter, iteratively optimize the clamping parameter, obtain an optimal clamping parameter, and perform assembly processing of the variable-diameter core sleeve, comprising: The clamping precision parameter and the clamping safety parameter are calculated based on the clamping influence parameter and the clamping damage parameter.
[0057] The assembly score is calculated based on the clamping precision parameter, the clamping safety parameter and the clamping polishing parameter.
[0058] The clamping parameter is randomly generated again, and the assembly score is calculated. Iterative optimization is performed until convergence, an optimal clamping parameter with the maximum assembly score is obtained, and assembly processing of the variable-diameter core sleeve is performed.
[0059] The variable-diameter assembly needs to simultaneously optimize the clamping precision, damage suppression and polishing efficiency, but the traditional method does not establish a multi-target collaborative mechanism, which may lead to a local optimal result, such as only achieving a low damage target but the polishing rate may be insufficient.
[0060] In the embodiments of the present application, the clamping accuracy parameter and the clamping safety parameter are calculated according to the clamping influence parameter and the clamping damage parameter. The clamping accuracy parameter = 1 - (clamping influence parameter ÷ preset clamping inner diameter), for example, the clamping influence parameter is 0.5 mm, and the preset clamping inner diameter is 10 mm, then the clamping accuracy parameter = 1 - (0.5 ÷ 10) = 0.95. The clamping safety parameter = 1 - (clamping damage parameter ÷ clamping front bobbin sleeve diameter), for example, the clamping damage parameter is 1 mm, and the clamping front bobbin sleeve diameter is 15 mm, then the clamping safety parameter = 1 - (1 ÷ 15) = 0.93.
[0061] The assembly score is calculated according to the clamping accuracy parameter, the clamping safety parameter and the clamping polishing parameter. The assembly score = clamping accuracy parameter + clamping safety parameter + clamping polishing parameter, for example, the clamping accuracy parameter is 0.95, the clamping safety parameter is 0.93, and the clamping polishing parameter is 0.8, then the assembly score = 0.95 + 0.93 + 0.8 = 2.68.
[0062] The clamping parameters and the assembly score are randomly generated again and iteratively optimized until convergence, for example, if the assembly score does not increase after 20 iterations, the clamping parameters corresponding to the maximum assembly score in the obtained assembly score are taken as the optimal clamping parameters, and the assembly processing of the variable-diameter bobbin sleeve is performed. Alternatively, a swarm intelligence algorithm such as a particle swarm algorithm can be used to optimize the clamping parameters, and the clamping parameters corresponding to the maximum assembly score after convergence are taken as the optimal clamping parameters.
[0063] The clamping accuracy parameter, the damage parameter and the polishing rate are unified and quantified by the assembly score function, and a multi-objective joint optimization is constructed. Based on the iteration mechanism, the optimal clamping parameters that balance the accuracy, safety and polishing are obtained. The obtained optimal clamping parameters can respond to the change of the inner diameter and ensure stable quality output in variable-diameter production.
[0064] Embodiment two, as shown in Figure 2 the same inventive concept as the intelligent assembly system for adaptive variable-diameter production of the bobbin sleeve provided in embodiment one, the present application embodiment further provides an intelligent assembly method for adaptive variable-diameter production of the bobbin sleeve, comprising: S10: During variable-diameter production of the bobbin sleeve, an image of the bobbin sleeve produced according to the changed inner diameter is collected, and a variable-diameter burr feature parameter is identified and obtained.
[0065] Among them, during the variable-diameter production of the bobbin sleeve, the image of the bobbin sleeve produced according to the changed inner diameter is collected, and the variable-diameter burr feature parameter is identified and obtained, comprising: During the variable-diameter production of the bobbin sleeve, the image of the bobbin sleeve produced according to the changed inner diameter is collected, and the variable-diameter image is obtained.
[0066] The variable-diameter image is input into the burr feature recognizer, and a variable-diameter burr feature parameter is obtained through identification output.
[0067] The training step of the burr feature recognizer includes: Based on the production data of the core sleeve in the historical time, a sample image set is collected, and burr feature parameters of the core sleeve in different sample images are collected to obtain a sample burr feature parameter set, wherein each sample burr feature parameter includes a burr size.
[0068] The burr feature recognizer is constructed based on deep learning.
[0069] The sample image set and the sample burr feature parameter set are used to iteratively train and optimize the burr feature recognizer, and the training is completed after the test loss meets the convergence requirement.
[0070] S20: Randomly generate clamping parameters for clamping and polishing of the variable-inner-diameter core sleeve, combine the variable-diameter burr feature parameters, and perform clamping influence analysis and clamping damage analysis to obtain clamping influence parameters and clamping damage parameters.
[0071] The randomly generated clamping parameters for clamping and polishing of the variable-inner-diameter core sleeve are combined with the variable-diameter burr feature parameters to perform clamping influence analysis and clamping damage analysis to obtain clamping influence parameters and clamping damage parameters, including: Randomly generate clamping parameters for clamping and polishing of the variable-inner-diameter core sleeve, wherein the clamping parameters include clamping force parameters.
[0072] The clamping parameters and the variable-diameter burr feature parameters are input into the clamping influence analysis branch and the clamping damage analysis branch to obtain clamping influence parameters and clamping damage parameters, wherein the clamping influence parameters include clamping influence size, and the clamping damage parameters include damage size.
[0073] The clamping influence analysis branch and the clamping damage analysis branch are trained through the following steps: According to the historical production data of the inner-diameter core sleeve, a sample clamping parameter set and a sample variable-diameter burr feature parameter set are collected, and size precision change values under clamping of different sample clamping parameters and sample variable-diameter burr feature parameters are collected to obtain a sample clamping influence parameter set. The damage size of the core sleeve after clamping under different sample clamping parameters and sample variable-diameter burr feature parameters is collected to obtain a sample clamping damage parameter set.
[0074] The clamping influence analysis branch and the clamping damage analysis branch based on machine learning are constructed.
[0075] Adopting the sample clamping parameter set and the sample variable-diameter burr feature parameter set as input data, respectively adopting the sample clamping influence parameter set and the sample clamping damage parameter set as output data, performing iterative training optimization on the clamping influence analysis branch and the clamping damage analysis branch respectively, and completing training after the test loss meets the convergence requirement.
[0076] S30: performing clamping polishing analysis according to the clamping parameter and the variable-diameter burr feature parameter, and obtaining a clamping polishing parameter.
[0077] The clamping polishing analysis according to the clamping parameter and the variable-diameter burr feature parameter to obtain the clamping polishing parameter includes: Obtaining a clamping polishing classifier, wherein the clamping polishing classifier is constructed by machine learning.
[0078] Inputting the clamping parameter and the variable-diameter burr feature parameter into the clamping polishing classifier to obtain a clamping polishing parameter through classification output, wherein the clamping polishing parameter includes a polishing rate.
[0079] S40: calculating an assembly score based on the clamping influence parameter, the clamping damage parameter and the clamping polishing parameter, iteratively optimizing the clamping parameter to obtain an optimal clamping parameter, and performing assembly processing of the variable-diameter shuttle core sleeve.
[0080] The calculation of the assembly score based on the clamping influence parameter, the clamping damage parameter and the clamping polishing parameter, the iterative optimization of the clamping parameter to obtain the optimal clamping parameter, and the assembly processing of the variable-diameter shuttle core sleeve include: According to the clamping influence parameter and the clamping damage parameter, clamping precision parameters and clamping safety parameters are calculated and obtained.
[0081] According to the clamping precision parameter, the clamping safety parameter and the clamping polishing parameter, an assembly score is calculated and obtained.
[0082] The clamping parameter is randomly generated again and the assembly score is calculated, and iterative optimization is performed until convergence, so as to obtain the optimal clamping parameter with the maximum assembly score, and perform assembly processing of the variable-diameter shuttle core sleeve.
[0083] In summary, the embodiments of the present application have at least the following technical effects: The application provides an intelligent assembly system and method for self-adaptive sizing production of a bobbin sleeve. By dynamically capturing the sizing burr features and constructing a collaborative optimization mechanism of clamping parameters-damage-polishing, the yield and size consistency of the sizing bobbin sleeve assembly are significantly improved. Compared with the traditional method, the technical scheme provided by the application significantly overcomes the adaptability defects of the fixed parameter assembly system in response to the sizing production: first, based on the burr feature analysis capability of real-time image recognition, the system can accurately perceive the burr shape difference of different inner diameter bobbin sleeves, laying the foundation for dynamic optimization; second, through the joint analysis model of clamping influence and damage, the damage caused by the mismatch of clamping force is considered in the parameter generation stage; third, the polishing efficiency is included in the multi-objective optimization function to avoid the problems such as insufficient polishing caused by the mismatch of the clamping parameters determined by the traditional method; finally, through the iteration mechanism of the assembly score, the clamping parameters are optimized automatically to realize the three-dimensional collaboration of clamping precision control, damage suppression and polishing efficiency, so as to maintain stable assembly quality in the sizing continuous production scene, and achieve the technical effects of self-adaptive adaptation of bobbin sleeves with various inner diameter specifications without manual intervention, synchronous improvement of assembly stability, safety and polishing precision.
[0084] It should be noted that the above sequence of the embodiments of the application is only for description, and does not represent the advantages and disadvantages of the embodiments. Moreover, the above describes specific embodiments of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.
[0085] The above only describes the preferred embodiments of the application and does not limit the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.
[0086] The present application and the drawings are only exemplary descriptions of the application, and any and all modifications, changes, combinations or equivalents within the scope of the application are considered to be covered by the present application. Obviously, those skilled in the art can make various modifications and changes to the application without departing from the scope of the application. Thus, if these modifications and changes of the application belong to the scope of the application and its equivalents, the application intends to include these modifications and changes.
Claims
1. An intelligent assembly system for adaptive variable diameter production of bobbin sleeves, characterized in that, The system includes: The image acquisition and recognition module is used to acquire images of the bobbin sleeve produced according to the changing inner diameter during the production of the bobbin sleeve, and to identify and obtain the characteristic parameters of the changing diameter burrs. The clamping damage influence analysis module is used to randomly generate clamping parameters for clamping and polishing the variable inner diameter shuttle core sleeve, and combine the variable diameter burr characteristic parameters to perform clamping influence analysis and clamping damage analysis to obtain clamping influence parameters and clamping damage parameters. The clamping and polishing analysis module is used to perform clamping and polishing analysis based on the clamping parameters and the variable diameter burr characteristic parameters to obtain the clamping and polishing parameters. The clamping parameter optimization module is used to calculate the assembly score based on the clamping influence parameters, clamping damage parameters, and clamping polishing parameters, iteratively optimize the clamping parameters, obtain the optimal clamping parameters, and perform the assembly and processing of the variable diameter shuttle sleeve.
2. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 1, characterized in that, During the production of bobbin sleeves with varying inner diameters, images of bobbin sleeves produced according to the changing inner diameter are captured, and characteristic parameters of the changing diameter burrs are identified, including: During the production of bobbin sleeves with varying inner diameters, images of bobbin sleeves produced according to the varying inner diameters are captured to obtain images of the varying diameters. The variable diameter image is input into the burr feature recognizer, and the recognition output obtains the variable diameter burr feature parameters.
3. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 2, characterized in that, The training steps for the burr feature recognizer include: Based on the production data of bobbin sleeves over a historical period, a set of sample images was collected, and the burr feature parameters of bobbin sleeves in different sample images were collected to obtain a set of sample burr feature parameters, wherein each sample burr feature parameter includes the burr scale. A burr feature recognizer was built based on deep learning. The spur feature recognizer is iteratively trained and optimized using the sample image set and the sample spur feature parameter set. Training is completed after the test loss meets the convergence requirement.
4. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 1, characterized in that, Clamping parameters for polishing a bobbin sleeve with a variable inner diameter are randomly generated. Combined with the characteristic parameters of the variable diameter burrs, clamping influence analysis and clamping damage analysis are performed to obtain clamping influence parameters and clamping damage parameters, including: Randomly generate clamping parameters for polishing a bobbin sleeve with a varying inner diameter, where the clamping parameters include clamping force parameters; The clamping parameters and the variable diameter burr characteristic parameters are input into the clamping influence analysis branch and the clamping damage analysis branch to obtain the clamping influence parameters and clamping damage parameters. The clamping influence parameters include the clamping influence dimension, and the clamping damage parameters include the damage dimension.
5. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 4, characterized in that, The clamping effect analysis branch and the clamping damage analysis branch are trained through the following steps: Based on historical production data of the same inner diameter bobbin sleeve, a set of sample clamping parameters and a set of sample diameter-changing burr characteristic parameters were collected. The dimensional accuracy variation values of clamping under different sample clamping parameters and sample diameter-changing burr characteristic parameters were also collected to obtain a set of sample clamping influence parameters. The damage dimensions of the spindle sleeve after clamping were collected under different sample clamping parameters and sample diameter variation burr characteristic parameters to obtain a set of sample clamping damage parameters. Construct a clamping effect analysis branch and a clamping damage analysis branch based on machine learning; Using the sample clamping parameter set and the sample variable diameter burr feature parameter set as input data, and the sample clamping influence parameter set and the sample clamping damage parameter set as output data, respectively, the clamping influence analysis branch and the clamping damage analysis branch are iteratively trained and optimized. The training is completed after the test loss meets the convergence requirement.
6. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 1, characterized in that, Based on the clamping parameters and the characteristic parameters of the variable diameter burr, a clamping and polishing analysis is performed to obtain the clamping and polishing parameters, including: Obtain a clamping and polishing classifier, wherein the clamping and polishing classifier is constructed using machine learning; The clamping parameters and variable diameter burr feature parameters are input into the clamping and polishing classifier, and the clamping and polishing parameters are obtained by classification output, wherein the clamping and polishing parameters include the polishing rate.
7. The intelligent assembly system for adaptive variable diameter production of bobbin sleeves according to claim 1, characterized in that, Based on the clamping influence parameters, clamping damage parameters, and clamping polishing parameters, the assembly score is calculated, the clamping parameters are iteratively optimized to obtain the optimal clamping parameters, and the assembly processing of the variable diameter bobbin sleeve is performed, including: Based on the clamping influence parameters and clamping damage parameters, the clamping accuracy parameters and clamping safety parameters are calculated. The assembly score is calculated based on the clamping accuracy parameters, clamping safety parameters, and clamping polishing parameters. The clamping parameters are randomly regenerated and the assembly score is calculated. Iterative optimization is performed until convergence, and the optimal clamping parameters with the largest assembly score are obtained. The variable diameter shuttle sleeve is then assembled.
8. An intelligent assembly method for adaptive variable diameter production of bobbin sleeves, characterized in that, The method includes: During the production of bobbin sleeves with variable inner diameter, images of bobbin sleeves produced according to the variable inner diameter are collected, and characteristic parameters of variable diameter burrs are identified. The clamping parameters for the polishing of the variable inner diameter bobbin sleeve are randomly generated. Combined with the variable diameter burr characteristic parameters, clamping influence analysis and clamping damage analysis are performed to obtain clamping influence parameters and clamping damage parameters. Based on the clamping parameters and the characteristic parameters of the variable diameter burr, a clamping and polishing analysis is performed to obtain the clamping and polishing parameters; Based on the clamping influence parameters, clamping damage parameters, and clamping polishing parameters, the assembly score is calculated, the clamping parameters are iteratively optimized, the optimal clamping parameters are obtained, and the variable diameter shuttle sleeve is assembled and processed.