Method for determining the engagement resistance of a waterproof zipper for luggage

Abstract

The application is suitable for the technical field of computer-aided zipper design, and particularly relates to a luggage waterproof zipper engagement resistance determination method, which comprises the following steps: acquiring actual test data and a three-dimensional model of a to-be-tested luggage waterproof zipper; constructing a simulation model of the to-be-tested luggage waterproof zipper based on the three-dimensional model; performing simulation test on the simulation model based on the test time and the test stroke to obtain a simulation result; iteratively optimizing key parameters of the simulation model based on the simulation result and the actual test data to obtain a calibrated simulation model; and performing simulation test on the calibrated simulation model based on the test time and the test stroke to obtain the engagement resistance of the to-be-tested luggage waterproof zipper. The method not only breaks through the limitation that traditional test methods can only obtain total resistance and errors caused by structural damage, but also significantly improves the fineness and accuracy of resistance analysis.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of computer-aided zipper design, and particularly relates to a luggage waterproof zipper engagement resistance determination method. BACKGROUND

[0002] The luggage waterproof zipper is usually equipped with a sealing strip, a coating or special waterproof material, which increases the friction of the zipper and may affect the smoothness and use experience of the zipper. The engagement resistance refers to the friction between the chain teeth and the slider and the chain teeth, which directly affects the sliding feeling and durability of the zipper. In order to ensure that the luggage waterproof zipper has waterproof performance while still having a smooth opening and closing experience, separately measuring the engagement resistance helps designers optimize the tooth shape, zipper head material and sealing strip design, thereby improving the service life, smoothness and durability of the zipper, meeting high-quality standards and consumer demand.

[0003] In the prior art, the performance evaluation method of the luggage waterproof zipper mainly relies on a physical testing machine to measure the total resistance of the zipper during the opening and closing process. The total resistance cannot reflect the real contribution of each part of the resistance, making it difficult for designers to quantitatively analyze and optimize the chain tooth structure, slider geometry or sealing strip material. Only relying on experience to repeatedly trial-produce, the development cycle is long and the cost is high.

[0004] Directly testing the engagement resistance by cutting off the sealing strip and other methods will damage the original structural characteristics of the luggage waterproof zipper, causing changes in the stress state of the chain tooth, the running track of the slider and the chain belt constraint condition, thereby making the values obtained by testing seriously deviate from the real engagement resistance.

[0005] In summary, when testing the engagement resistance of the luggage waterproof zipper, there is a problem that the limitation of the testing method leads to the inability to accurately obtain the engagement resistance. SUMMARY

[0006] The embodiments of the application provide a luggage waterproof zipper engagement resistance determination method, which can solve the problem in the related art that when testing the engagement resistance of the luggage waterproof zipper, the limitation of the testing method leads to the inability to accurately obtain the engagement resistance.

[0007] In a first aspect, the embodiments of the application provide a luggage waterproof zipper engagement resistance determination method, including:

[0008] Obtaining actual test data and a three-dimensional model of a to-be-tested luggage waterproof zipper; wherein the actual test data includes total resistance-displacement curves under four environmental conditions, test time and test stroke, and the four environmental conditions include normal temperature dry state, normal temperature wet state, low temperature dry state and low temperature wet state, and the total resistance includes engagement resistance and sealing resistance;

[0009] constructing a simulation model of the waterproof zipper to be tested based on the three-dimensional model; wherein the simulation model comprises key parameters and contact pairs, the key parameters comprise super-elasticity parameters of the sealing strips, friction coefficients between the sealing strips, friction coefficients between the chain teeth, four groups of key parameters corresponding to the four environmental conditions, the contact pairs comprise meshing contact pairs and sealing contact pairs, the meshing contact pairs are used to simulate mechanical meshing interactions between the chain teeth and the chain teeth and between the chain teeth and the puller, and the sealing contact pairs are used to simulate compression sealing interactions between the sealing strips and the sealing strips;

[0010] performing simulation testing on the simulation model based on the test time and the test stroke to obtain simulation results; wherein the simulation results comprise total resistance-displacement curves, meshing resistance-displacement curves and sealing resistance-displacement curves under the four environmental conditions;

[0011] performing iteration optimization on the key parameters of the simulation model based on the simulation results and the actual test data to obtain a calibrated simulation model;

[0012] performing simulation testing on the calibrated simulation model based on the test time and the test stroke to obtain meshing resistance of the waterproof zipper to be tested.

[0013] The technical solutions described above in the embodiments of the present application have at least the following technical effects:

[0014] The method for determining the engagement resistance of a luggage waterproof zipper provided in the application comprises the following steps: obtaining actual test data and a three-dimensional model of a to-be-tested luggage waterproof zipper, so as to obtain total resistance-displacement curves under four environmental conditions of normal temperature dry state, normal temperature wet state, low temperature dry state and low temperature wet state as well as test time and test stroke; constructing a simulation model of the to-be-tested luggage waterproof zipper based on the three-dimensional model, so as to simulate the super-elasticity parameters of the sealing strip of the actual situation luggage waterproof zipper, the friction coefficient between the sealing strips, the friction coefficient between the chain teeth, the mechanical engagement interaction (engagement resistance) between the chain teeth and the slider and the compression sealing interaction (sealing resistance) between the sealing strips; performing simulation test on the simulation model based on the test time and the test stroke, so as to obtain simulation results, and to obtain the total resistance-displacement curves, the engagement resistance-displacement curves and the sealing resistance-displacement curves under the four environmental conditions; iteratively optimizing the key parameters of the simulation model based on the simulation results and the actual test data, so as to obtain a calibrated simulation model; and performing simulation test on the calibrated simulation model based on the test time and the test stroke, and extracting the engagement resistance-displacement curves of the to-be-tested luggage waterproof zipper from the simulation results. The method can independently define the engagement contact pairs and the sealing contact pairs in the simulation, iteratively optimize the key parameters in combination with the measured total resistance-displacement curves, and construct a simulation model which can quantitatively predict the contribution proportion of the two types of resistance at different displacement positions on the premise of maintaining the integrity of the luggage waterproof zipper structure. The method not only breaks through the limitation of the traditional test method that can only obtain the total resistance and the error caused by the structure damage, but also significantly improves the fineness and accuracy of the resistance analysis, so that the designers can directly obtain the real engagement resistance distribution based on the calibrated simulation results, thereby replacing a large number of traditional trial and detection processes. The method can establish a parameterized simulation model of four working conditions, can truly reflect the influence of temperature and humidity on the material properties and friction behavior of the sealing strip, can make the simulation results highly consistent with the actual use environment, can significantly expand the test dimension, and can enhance the comprehensiveness and reliability of the zipper performance evaluation.

[0015] In a second aspect, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the method of any of the embodiments in the first aspect when executing the computer program.

[0016] It can be understood that the beneficial effects of the above-mentioned second aspect can be referred to the related description in the above-mentioned first aspect, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0018] Figure 1 is a flowchart of a method for determining the engagement resistance of a waterproof zipper of a luggage bag according to an embodiment of the present application;

[0019] Figure 2 is a comparison chart of the actual total resistance and the simulated total resistance, the simulated engagement resistance and the simulated sealing resistance under normal temperature dry state in the method for determining the engagement resistance of a waterproof zipper of a luggage bag according to an embodiment of the present application;

[0020] Figure 3 is a comparison chart of the actual total resistance and the simulated total resistance, the simulated engagement resistance and the simulated sealing resistance under low temperature dry state in the method for determining the engagement resistance of a waterproof zipper of a luggage bag according to an embodiment of the present application;

[0021] Figure 4 is a comparison chart of the actual total resistance and the simulated total resistance, the simulated engagement resistance and the simulated sealing resistance under normal temperature wet state in the method for determining the engagement resistance of a waterproof zipper of a luggage bag according to an embodiment of the present application;

[0022] Figure 5 is a comparison chart of the actual total resistance and the simulated total resistance, the simulated engagement resistance and the simulated sealing resistance under low temperature wet state in the method for determining the engagement resistance of a waterproof zipper of a luggage bag according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] In the following description, for the purpose of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0024] It should be understood that when used in the specification and the appended claims, the term "comprise" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025] It should also be understood that the term "and / or" as used herein in the specification and in the claims, unless otherwise specified, means any one of the associated listed items or a combination of any of the associated listed items.

[0026] In the related art, the performance evaluation method of the luggage waterproof zipper mainly depends on the physical testing machine to measure the resistance of the zipper during opening and closing. However, such testing means has obvious limitations. The traditional testing means can only obtain the total resistance of the puller during the entire opening or closing stroke, which is the superposition result of the meshing resistance and the sealing resistance. For the luggage waterproof zipper, the mechanical meshing action between the chain teeth and the puller will generate meshing resistance, and the compression contact between the sealing strips, material deformation and interface friction will cause sealing resistance. Since the existing testing means cannot separate the two mechanisms, the total resistance obtained cannot reflect the true contribution of each part of the resistance, making it difficult for designers to determine whether the resistance abnormality comes from the meshing structure or the sealing structure, and lacking basis for targeted optimization of the chain tooth shape, puller structure or sealing strip material.

[0027] The existing tests are mainly carried out under normal temperature and dry state conditions, and can only reflect the performance of the zipper in a single environment, while the luggage waterproof zipper often faces different working conditions (such as humid environment) in actual use. The temperature and humidity environment will significantly affect the mechanical properties and interface friction behavior of the sealing strip material, causing changes in meshing resistance and sealing resistance. In the absence of multi-environment test data, it is difficult to comprehensively evaluate the actual service performance of the waterproof zipper, and it is also difficult to provide effective reference for structure design and material selection under different use environments.

[0028] In addition, since the test results only include the total resistance and the environment coverage is insufficient, the existing technology cannot directly and efficiently obtain the meshing resistance itself. Even if it tries to reduce the influence of sealing resistance by cutting off the sealing strip, it will also damage the original structural characteristics of the zipper, causing changes in the stress state of the chain tooth, the running track of the puller and the chain belt constraint condition, so that the values obtained by the test are seriously deviated from the true meshing resistance. The inability to accurately obtain the meshing resistance means that designers are difficult to carry out quantitative analysis and optimization on the chain tooth structure, puller geometry or meshing cooperation relationship, and can only rely on experience to repeatedly trial-produce, which has long research and development cycle and high cost.

[0029] To address the aforementioned issues, this application provides a method for determining the engagement resistance of waterproof zippers for bags. This method independently defines the engagement contact pair and the sealing contact pair in the simulation, and iteratively optimizes key parameters by combining the measured total resistance-displacement curve. The resulting simulation model can quantitatively predict the contribution ratio of the two types of resistance at different displacement positions while maintaining the structural integrity of the waterproof zipper. This not only overcomes the limitations of traditional testing methods that can only obtain total resistance and errors caused by structural damage, but also significantly improves the precision and accuracy of resistance analysis. It allows designers to directly obtain the true engagement resistance distribution based on the calibrated simulation results, thus replacing a large amount of traditional trial and testing processes. By establishing parameterized simulation models for four operating conditions, this method can realistically reflect the influence of temperature and humidity on the material properties and frictional behavior of the sealing strip, making the simulation results highly consistent with the actual usage environment. This significantly expands the testing dimensions and enhances the comprehensiveness and reliability of zipper performance evaluation.

[0030] The method for determining the engagement resistance of waterproof zippers for bags provided in this application can be applied to electronic devices. In this case, the electronic device is the executing subject of the method for determining the engagement resistance of waterproof zippers for bags provided in this application. This application does not impose any restrictions on the specific type of electronic device.

[0031] For example, electronic devices can be tablets, laptops, ultra-mobile personal computers (UMPCs), desktop computers, computers, laptops, etc.

[0032] To better understand the method for determining the engagement resistance of waterproof zippers for bags provided in this application, the specific implementation process of the method for determining the engagement resistance of waterproof zippers for bags provided in this application will be described by way of example below.

[0033] Figure 1 This paper presents a schematic flowchart illustrating a method for determining the engagement resistance of a waterproof zipper for bags, as provided in an embodiment of this application. The method includes:

[0034] S100 acquires the actual test data and 3D model of the waterproof zipper of the bag under test. The actual test data includes the total resistance-displacement curve, test time, and test stroke under four environmental conditions: normal temperature dry state, normal temperature humid state, low temperature dry state, and low temperature humid state. The total resistance includes engagement resistance and sealing resistance.

[0035] It can be understood that the total resistance is the overall resistance experienced by the puller during the opening and closing of the zipper, including the mechanical resistance (meshing resistance) generated when the chain teeth mesh with each other, and the chain teeth and the puller slide and extrude relative to each other; the resistance (sealing resistance) generated during the compression, contact and relative sliding of the sealing strip.

[0036] The total resistance-displacement curve represents the change of resistance during the displacement of the puller from the starting point to the ending point; the test time represents the total time experienced from the start of the puller movement to the completion of the test stroke at a uniform pulling speed; the test stroke represents the displacement distance experienced by the puller during the test.

[0037] Normal temperature dry state environmental conditions (such as 23°C, 50% RH) are used to simulate normal temperature and dry environment, in which the zipper is not affected by external moisture; normal temperature wet state environmental conditions are used to simulate normal temperature and humid environment, which is used to simulate the influence of rain or humid environment on the waterproof zipper of luggage; low temperature dry state environmental conditions are used to simulate low temperature and dry environment (such as -10°C, 50% RH); low temperature wet state environmental conditions are used to simulate low temperature and humid environment.

[0038] Exemplarily, not less than 5 perfect waterproof zippers of luggage can be randomly selected from the same production batch as the waterproof zipper of luggage to be tested as test samples. An electronic tensile testing machine equipped with displacement sensors (for recording the displacement of the puller) and force sensors (for measuring the total resistance generated during opening and closing) can be used; an environmental control system can be used to adjust the temperature and humidity of the test environment in order to simulate different environmental conditions (such as an environmental simulation test box).

[0039] Normal temperature dry state: the temperature in the test box can be set to normal temperature (such as 23°C) and the humidity can be set to a conventional dry level (such as 50% RH); normal temperature wet state: the temperature in the test box can be maintained at normal temperature while increasing the humidity (such as 90% RH); low temperature dry state: the temperature in the test box can be adjusted to low temperature (such as -10°C) while maintaining the humidity at a conventional level (such as 50% RH); low temperature wet state: the temperature can be maintained at low temperature (such as -10°C) while increasing the humidity (such as 90% RH).

[0040] Through the above environmental settings, the opening and closing test is carried out under each environmental condition: the test speed and test stroke can be set in the control software of the electronic tensile testing machine (consistent in the four environmental conditions), one end of the fabric tape of the test sample can be placed in the lower clamp, the clamping position can be about 10-15 mm away from the edge of the chain teeth, the fabric tape position can be adjusted so that the center line of the zipper is in the same vertical plane as the loading direction of the testing machine, to avoid generating torsion or lateral force; the hook part of the upper clamp or a special clamping structure can be used to securely hang the puller tab, and a slight pre-tension (such as 0.5-1 N) can be applied to eliminate system gaps, which is beneficial to the straightening of the zipper fabric tape and chain teeth at the beginning without folding.

[0041] Start the test machine, the upper clamp moves at a constant test speed, driving the puller to move along the chain tooth direction, completing the process of opening to closing the zipper, the displacement sensor and the force sensor synchronously collect the displacement data and the total resistance at each time point, and generate the total resistance-displacement curve in the control software, while recording the test time. The same sample under each environmental condition can be tested at least three times to eliminate possible accidental errors.

[0042] Through the above test process, the total resistance-displacement curve, test time, and test stroke under four environmental conditions can be obtained.

[0043] The complete zipper assembly CAD model (three-dimensional model of the test sample) including chain teeth, puller, tape, and waterproof sealing strip can be obtained from the zipper supplier or internal design department; three-dimensional scanning technology (such as laser scanner, optical three-dimensional scanner, or structured light scanner) can be used to scan and model the test sample, generating point cloud data, which is converted into an accurate three-dimensional model (such as STL or OBJ format) through data processing and reconstruction algorithms.

[0044] The actual test data provide a basis for further analysis of the engagement and sealing performance of the waterproof zipper of the luggage, and the three-dimensional model provides necessary geometric data for subsequent simulation analysis. Through the above steps, the performance of the waterproof zipper under different environments can be more comprehensively understood, and data support is provided for subsequent simulation and optimization.

[0045] S200, based on the three-dimensional model, a simulation model of the waterproof zipper of the luggage to be tested is constructed. The simulation model includes key parameters and contact pairs, the key parameters include the hyperelastic parameters of the sealing strip, the friction coefficient between the sealing strips, and the friction coefficient between the chain teeth, four sets of key parameters corresponding to the four environmental conditions, the contact pairs include the engagement contact pairs and the sealing contact pairs, the engagement contact pairs are used to simulate the mechanical engagement interaction between the chain teeth and the chain teeth, and the chain teeth and the puller, and the sealing contact pairs are used to simulate the compression sealing interaction between the sealing strips and the sealing strips.

[0046] It can be understood that the sealing strip is made of hyperelastic material (such as silicone, rubber, etc.), which has nonlinear elastic properties and can restore to its original shape after being deformed. The behavior of the hyperelastic material can be described using a hyperelastic model, such as the Mooney-Rivlin model. C10 and C01 are two hyperelastic parameters in the Mooney-Rivlin model, C10 is related to the elastic behavior of the material, indicating the stress-strain relationship of the material during stretching or compression; C01 is related to the nonlinear behavior of the material, describing the flexibility or rigidity of the material under large deformation.

[0047] The coefficient of friction between sealing strips determines the magnitude of frictional resistance during the contact process between the sealing strips.

[0048] The coefficient of friction between chain teeth determines the magnitude of the resistance during the meshing process between chain teeth and between chain teeth and the pull head.

[0049] For example, the 3D model can be imported into simulation software (such as ANSYS, ABAQUS, COMSOL, etc.), which supports finite element analysis (FEA) of complex geometries and can simulate and predict the working performance of zippers in different environments through numerical methods.

[0050] The model can be divided into independent geometric components such as chain teeth, pull heads, and sealing strips, which facilitates the subsequent definition of material properties and contact relationships; local details can be appropriately simplified while retaining the key areas that have the most significant impact on mechanical response; and potential geometric gaps, overlapping surfaces, or discontinuities can be corrected, which helps the model to successfully complete mesh generation.

[0051] After completing the above steps, the structure can be meshed, and a higher density mesh arrangement can be made for the chain tooth meshing area and the sealing strip contact area in order to more accurately capture contact changes and local mechanical responses.

[0052] All contact surfaces of the zipper teeth in the model can be defined as contactable surfaces; a contact algorithm based on the penalty function or the extended Lagrange multiplier method can be used to simulate the actual meshing process; the friction coefficient between the zipper teeth can be set to ensure that the frictional resistance during meshing is accurately reflected. The meshing contact pairs defined through these steps can realistically reproduce the interaction during zipper meshing.

[0053] The contact surfaces of the sealing strips on the left and right sides of the model can be defined as a pair of contact surfaces; the contact properties between the sealing strips can be set, including normal hard contact and tangential frictional contact; the friction coefficient between the sealing strips can be set so that the friction force generated during the compression deformation of the sealing strip can be accurately simulated; the elastic deformation of the sealing strip can be defined by setting hyperelastic parameters (C10, C01) so that the relationship between compression and reaction force in the sealing contact is as close as possible to reality.

[0054] Four sets of parameters can be established in the simulation to simulate the state of the waterproof zipper of the bag under test under four environmental conditions, including C10, C01 at room temperature, friction coefficient between sealing strips in dry state, friction coefficient between chain teeth in dry state; C10, C01 at room temperature, friction coefficient between sealing strips in wet state, friction coefficient between chain teeth in wet state; C10, C01 at low temperature, friction coefficient between sealing strips in dry state, friction coefficient between chain teeth in dry state; and C10, C01 at low temperature, friction coefficient between sealing strips in wet state, friction coefficient between chain teeth in wet state.

[0055] Corresponding boundary conditions and simulation loads can be established in the simulation, including fixing one end of the zipper as a constraint reference, setting the puller or the chain tape on one side as a movable component, applying a displacement load in the simulation to move the puller along the actual opening and closing path, and setting the displacement amount as the test stroke.

[0056] The simulation model is completed through the above process, and corresponding parameter sets can be called respectively to simulate four environmental conditions in the simulation process.

[0057] The simulation model constructed through the above steps can realistically simulate the mechanical behavior in the chain tooth meshing and sealing strip compression process, and provide an accurate and reliable basis for subsequent simulation testing, parameter calibration, and meshing resistance prediction.

[0058] In one possible implementation, S200, based on the three-dimensional model, a simulation model of the waterproof zipper to be tested is constructed, including:

[0059] S210, geometric cleaning and geometric simplification are performed on the three-dimensional model to obtain a first model.

[0060] Exemplarily, geometric cleaning: the simulation software can be used to check whether the three-dimensional model has problems such as broken surfaces, self-intersection, small gaps, and overlapping bodies; the detected open boundaries can be closed using tools such as merging edges / closing surfaces; small gaps (much smaller than the feature size and having negligible effect on contact behavior) can be merged into continuous entities using the merge / stitch function; overlapping bodies can be merged or trimmed using Boolean operations (such as Unite or Overlap) to avoid multiple bodies occupying the same space.

[0061] Geometric simplification: small decorative logos and concave patterns on the puller that have little effect on the mechanical behavior can be deleted; very small fillets and chamfers (such as <0.1 mm and far from the contact area) can be removed using the chamfer / round corner removal function to simplify them into sharp edges or large curvature surfaces; the surface texture of the fabric tape and the stitches can not be modeled as entities, and the fabric tape can be equivalent to a uniform thickness entity plate or shell unit geometry, retaining the overall profile size (length, width, thickness) of the fabric tape, which is beneficial to the consistency of its macroscopic stiffness with the actual situation; the real contact area geometry of the sealing strip and the chain tooth, the puller, and the sealing strip on the opposite side can be retained without excessive simplification to facilitate accurate simulation of sealing contact and meshing behavior.

[0062] Through the above process, a clean, non-leakage, and topologically clear geometric model (first model) can be obtained, providing a reliable basis for subsequent meshing and contact definition.

[0063] S220, meshing and local mesh refinement are performed on the first model to obtain a second model.

[0064] Exemplarily, for the tape and the sealing strip in the first model, if the thickness is small and the deformation is relatively flat, shell elements or thin solid elements can be used; if it is necessary to accurately capture the compression and contact deformation of the sealing strip in the thickness direction, three-dimensional solid elements (Hex or Tet) can be used.

[0065] For the sprocket and the puller in the first model, tetrahedral elements (Tet) can be used for automatic division, taking into account the geometric adaptability and mesh quality.

[0066] A global starting mesh size (such as 0.5-1.0 mm) can be set, and local mesh refinement can be performed in the sprocket meshing area (the position of the slider and the sprocket contact and disengagement), the contact passage area when the sprocket passes through the puller, the mutual contact area of the sealing strip, and the contact area of the sealing strip and the sprocket, the puller. The mesh size can be reduced to 1 / 2-1 / 4 of the global starting mesh size. Local mesh control or surface / body local division function can be used to divide the above-mentioned areas into finer meshes.

[0067] The aspect ratio, twist, Jacobian, and internal angle of the mesh elements can be checked, and the elements with obvious unqualified quality (such as severely twisted and inverted elements) can be re-divided or locally adjusted.

[0068] S230, material properties are added to the second model, and contact pairs and boundary conditions of the second model are defined to obtain a simulation model.

[0069] Exemplarily, in the material module of the simulation software, material properties can be established for different components: the material model of the puller and the sprocket in the second model can use a linear elastic material model, if the sprocket and the puller are nylon materials, the Young's modulus of the material model can be set to 2000-3000 MPa; the Poisson's ratio can be set to 0.3-0.4; the density can be input according to the actual value of the material; if the sprocket and the puller are metals (such as copper), the Young's modulus of the material model can be set to 110 GPa; the Poisson's ratio can be set to 0.33; the density can be defined according to the actual metal material.

[0070] The material model of the belt in the second model can adopt an orthotropic elastic or an engineered anisotropic material model to simulate the different mechanical responses of the fabric in the warp, weft and thickness directions. The parameter settings of the material model of the belt can be obtained through fabric tensile tests (such as the warp elastic modulus E1, the weft elastic modulus E2, the corresponding Poisson's ratios v12, v23, v31, the shear moduli G12, G23, G31); the material and thickness can be set in the section attribute of the belt. For example, the material of the belt is polyester, and the thickness is 0.32 mm, and the material model parameters thereof are: E1 is 3500 MPa, E2 is 1800 MPa, v12 is 0.28, v23 is 0.06, v31 is 0.10, G12 is 22 MPa, G23 is 6 MPa, and G31 is 6 MPa.

[0071] The material model of the sealing strip in the second model can adopt a hyperelastic constitutive model, such as Mooney-Rivlin, Yeoh, etc. Here, the Mooney-Rivlin model is used, and the uniaxial tensile test of the sealing strip at room temperature and low temperature can be performed to obtain the stress-strain curves at room temperature and low temperature, and the Mooney-Rivlin model is used to fit the two curves respectively (such as the least square method) to obtain the initial values of C10 and C01 at room temperature and low temperature, such as C10 at room temperature 23℃ is 0.8, and C01 is 0.2 (relatively soft); C10 at low temperature-10℃ is 2.4, and C01 is 0.6 (hard at low temperature).

[0072] Through the above process, the material properties of each component in the second model are clearly distinguished in the simulation.

[0073] The master surface of the mesh contact pair can be set as the guide channel surface in contact with the sprocket inside the puller, and the slave surface can be set as the outer surface of the sprocket. The contact properties of the mesh contact pair can include normal behavior (i.e. hard contact, the contact between components can be defined using non-penetrating contact), and tangential behavior (i.e. friction behavior on the contact surface, which can be defined by setting the friction coefficient between the sprockets).

[0074] The master and slave surfaces of the sealing contact pair can be set as the opposite contact surfaces of the left and right sealing strips. The contact properties of the sealing contact pair can include normal behavior (hard contact or softened contact (selected according to material properties)), and tangential behavior (i.e. friction behavior between the sealing strips, which can be defined by setting the friction coefficient between the sealing strips).

[0075] If the simulation software provides General Contact, all contact pairs can be put into the General Contact, and different contact property groups can be defined for different contact areas; if the Contact Pair method is used, independent contact pairs can be set for meshing contact and sealing contact, which is convenient for subsequent output of the contact force of each contact pair.

[0076] The test sample can be separated into a seal strip part and a sprocket part, and a friction tester is used to test the seal strip part and the sprocket part under dry and wet conditions according to a standard (such as ASTM D1894), and the initial values of the friction coefficients of the seal strip part and the sprocket part under dry and wet conditions are measured. For example, the inter-seal strip friction coefficient under dry conditions is 0.35, the inter-seal strip friction coefficient under wet conditions is 0.18, the inter-sprocket friction coefficient under dry conditions is 0.25, and the inter-sprocket friction coefficient under wet conditions is 0.12.

[0077] Temperature mainly affects the molecular motion inside the material, thereby affecting the macroscopic mechanical properties (such as modulus) of the material, and the influence on the friction coefficient is relatively secondary; and dry and wet conditions mainly change the lubrication state of the contact surface, thereby affecting the friction coefficient, and the influence on the constitutive behavior of the material is relatively secondary. Therefore, in the simulation model of the present application, only the influence of temperature on the super-elastic parameters (C10, C01) and the influence of dry and wet conditions on the friction coefficient can be considered, and through the above simplification, the complexity of the model can be reduced, and in actual application, the accuracy requirement can be met.

[0078] By clearly distinguishing meshing contact and sealing contact in simulation, meshing resistance and sealing resistance can be separately extracted from the results after solving.

[0079] The full constraint (all translational and rotational degrees of freedom are fixed) can be applied to the end of the belt of the second model to simulate the fixation of the lower clamp in the actual test. A reference point can be defined on the puller of the second model, and the puller is connected to the reference point through a multi-point constraint (such as MPC, Rigid Body), and a displacement boundary condition consistent with the test stroke in the sprocket direction is applied to the reference point.

[0080] Output settings can be defined in the simulation software: the reaction force of the reference point of the puller can be set for historical output, and the component in the pulling direction is the total resistance-displacement curve in the simulation; the meshing resistance-displacement curve can be requested for the meshing contact pair, such as through the Contact Force History function of the simulation software; and the sealing resistance-displacement curve can also be requested for the sealing contact pair.

[0081] By defining output settings for different contact pairs, basic data can be provided for subsequent comparison with the measured total resistance and parameter optimization.

[0082] S300, based on the test time and the test stroke, performing simulation test on the simulation model to obtain a simulation result. The simulation result includes total resistance-displacement curves, meshing resistance-displacement curves and sealing resistance-displacement curves under four environmental conditions.

[0083] Exemplarily, the simulation is performed at discrete time steps, and the size of the time step determines the accuracy and stability of each calculation. The test time can be set as the total analysis step time. The total analysis step time and the simulation displacement under each environmental condition are consistent, and are the same as the test time and the test stroke, so as to facilitate the connection of the time and displacement range of the simulation with the actual test.

[0084] According to the total analysis step time and the simulation displacement, the displacement of the puller of the simulation model is gradually advanced, the meshing contact pairs are solved in each time step, the friction and meshing compression force between the chain teeth are calculated to obtain the meshing resistance; the compression force and friction of the sealing contact pairs are calculated in each time step to obtain the sealing resistance; the contact force of each contact pair is updated in each time step, and the total resistance is obtained according to the sum of the contact forces. The changes of the contact force and the total resistance with the displacement are accumulated to generate the displacement curves of the total resistance, the meshing resistance and the sealing resistance. Four sets of parameter groups can be loaded into the simulation model, and the total resistance-displacement curves, the meshing resistance-displacement curves and the sealing resistance-displacement curves under four environmental conditions can be generated through the above simulation process.

[0085] The simulation result obtained through the step not only intuitively presents the meshing resistance characteristics of the simulation model under different environmental conditions, but also provides a complete basis for subsequent comparison with real-time test data, parameter calibration and evaluation of the relative proportion of the resistance contribution of the sealing strip and the chain teeth.

[0086] In a possible implementation, S300, based on the test time and the test stroke, performing simulation test on the simulation model to obtain a simulation result, including:

[0087] S310, constructing a displacement function according to the test time and the test stroke.

[0088] Exemplarily, a dynamic explicit analysis step can be created in the simulation software. The dynamic explicit solver (such as LS-DYNA and Abaqus Explicit) is good at handling complex contact and large deformation problems, and can more stably calculate the process of meshing in and out of a single chain tooth, and is not prone to calculation interruption due to local non-convergence.

[0089] The total analysis step time can be set as the test time, and the displacement boundary condition of the puller can be set as the test stroke, that is, the distance from the initial position (0) to the terminal position (L) of the reference point of the puller.

[0090] In explicit dynamic simulation, the displacement function can be assumed to be linearly varying, which can be expressed as , where T represents the total analysis step time.

[0091] The solver in the simulation software can scan the entire simulation model, and for each grid cell, according to its characteristic length and material wave speed, calculate the maximum stable time increment allowed for the cell, and take the minimum time increment among all grid cells as the time increment in the simulation process . The total analysis step time T is divided into n time steps according to the time increment , and the time advancement of one step can be expressed as For each time step , the solver can call the displacement function to calculate the displacement of the reference point of the puller at the current time step.

[0092] In the simulation process, the displacement function not only gives the reference displacement of the puller, but also helps the solver to determine the position and motion of each node at each time step. Through the displacement function, the displacement of the reference point of the puller at each time step is precisely controlled, and the behavior of the puller in the simulation can be simulated.

[0093] S320, for each environmental condition:

[0094] Based on the displacement function and the key parameters under the current environmental condition, the solver is used to solve the simulation model to obtain the total resistance-displacement curve, the meshing resistance-displacement curve, and the sealing resistance-displacement curve under the current environmental condition.

[0095] Exemplarily, for each time step: the solver can calculate the deformation of the sprocket and the puller through a linear elastic material model, and based on the deformation and the displacement at the current time step, calculate the normal force through a contact mechanics model (such as the Hertz contact model); according to the normal force and the friction coefficient between the sprocket and the puller, the tangential force is calculated, that is , where represents the friction coefficient between the sprocket and the puller. The sum of the normal force and the tangential force is the meshing resistance.

[0096] The solver can check the contact state between the sealing strips and calculate the contact area between the sealing strips at the current time step; through C10 and C01 in the Mooney-Rivlin model, the stress at the current time step is calculated, and the product of the stress and the contact area is calculated to obtain the normal force ; according to the normal force and the friction coefficient between the sealing strips, the tangential force is calculated, that is , where The sealing strip represents the friction coefficient between the sealing strip. The sum of the normal force and the tangential force is the sealing resistance.

[0097] The solver can calculate the strain and stress of each grid element through the material model of the belt according to the displacement of the current time step, and convert the stress to the equivalent internal force of the current time step through the element shape function. The sum of the internal force, the engagement resistance and the sealing resistance is the total resistance.

[0098] The solver can correspond each time step of the total resistance, the engagement resistance and the sealing resistance calculated by the above process to each time step of the displacement respectively, and output the total resistance-displacement curve, the engagement resistance-displacement curve and the sealing resistance-displacement curve under the current environmental condition after the simulation ends through the output setting.

[0099] Through the above steps, the solver can accurately simulate the mechanical behavior of the zipper system under different environmental conditions, and the three curves output reflect the force change of the zipper under different environmental conditions, which provides key mechanical data for analyzing the performance of the zipper and helps to optimize the design and verify the working state.

[0100] S400, based on the simulation results and the actual test data, iteratively optimizing the key parameters of the simulation model to obtain the calibrated simulation model.

[0101] Exemplarily, for each environmental condition: the loss function can be used: , wherein, represents the displacement of the total resistance-displacement curve of the simulation result , the corresponding total resistance, represents the displacement of the total resistance-displacement curve of the actual test data , the corresponding total resistance, N represents the number of data points in the total resistance-displacement curve, and the error between the total resistance-displacement curve of the simulation result and the total resistance-displacement curve of the actual test data is calculated.

[0102] The key parameters under the current environmental condition can be constructed as a parameter vector .

[0103] The optimization algorithm can select a gradient-based optimization method (such as quasi-Newton method, gradient descent method), or a derivative-free method (such as genetic algorithm, particle swarm algorithm or adaptive parameter search algorithm). If the gradient descent method is selected, the parameter vector can be updated by the formula: (wherein, represents the learning rate for controlling the step size of each update), and if the derivative-free method is selected, a new parameter vector can be generated according to the value of the loss function, crossover, mutation or velocity update rule.

[0104] The four updated parameter vectors under the four environmental conditions can be obtained through the above steps. Based on the four updated parameter vectors, the simulation model is simulated again, and the total resistance-displacement curve of the new simulation result is compared with the total resistance-displacement curve of the actual test data to calculate the error. The parameter vector updating, simulation testing, and error calculation process are repeatedly executed until the convergence condition is met (such as the loss function is reduced to below the set threshold value, or the maximum iteration number is reached), and finally the calibrated key parameters under each environmental condition are obtained.

[0105] This step minimizes the total resistance error between simulation and actual test data, so that the simulation model can accurately predict the sealing performance and friction change of the luggage waterproof zipper under different environmental conditions, which is beneficial to the design of the zipper to meet the performance requirements in the real use environment, and finally realizes the optimization of the reliability and durability of the product.

[0106] In one possible implementation, S400, based on the simulation results and the actual test data, the key parameters of the simulation model are iteratively optimized to obtain a calibrated simulation model, including:

[0107] S410, error calculation: according to the objective function, the simulation error between the total resistance-displacement curve in the simulation result and the total resistance-displacement curve in the actual test data is calculated.

[0108] Exemplarily, the simulation error between the simulation result and the actual test data for all environmental conditions can be calculated using the objective function, that is, where F represents the simulation error, represents the displacement corresponding to the total resistance in the total resistance-displacement curve under the jth environmental condition in the simulation result, represents the displacement corresponding to the total resistance in the total resistance-displacement curve under the jth environmental condition in the actual test data, and M represents the total number of environmental conditions.

[0109] Optionally, S410, according to the objective function, the simulation error between the total resistance-displacement curve in the simulation result and the total resistance-displacement curve in the actual test data is calculated, including:

[0110] For each environmental condition:

[0111] S411, according to each data point of the simulation total resistance-displacement curve and each data point of the measured total resistance-displacement curve under the current environmental condition, the root mean square error is calculated. The simulation total resistance-displacement curve is the total resistance-displacement curve in the simulation result, and the measured total resistance-displacement curve is the total resistance-displacement curve in the actual test data.

[0112] Exemplarily, the calculation formula of the root mean square error (RMSE) is The resistance corresponding to each displacement in the simulation total resistance-displacement curve can be substituted into the RMSE calculation formula as the resistance corresponding to each displacement in the measured total resistance-displacement curve, and the root mean square error can be obtained. The RMSE is used to measure whether the overall shape of the simulation curve is consistent with the measured curve.

[0113] S412, the simulation peak resistance and the actual peak resistance are extracted from the simulation total resistance-displacement curve and the measured total resistance-displacement curve of the current environmental condition respectively, and the peak error between the simulation peak resistance and the actual peak resistance is calculated.

[0114] Exemplarily, the peak resistances can be extracted from the simulation total resistance-displacement curve and the actual test total resistance-displacement curve respectively, and the absolute error between the two is calculated, that is, , wherein, represents the peak resistance in the simulation total resistance-displacement curve, represents the peak resistance of the measured total resistance-displacement curve, and P represents the peak error. The peak error can reflect the performance of the simulation model in the limit case.

[0115] S413, according to the simulation total resistance-displacement curve and the measured total resistance-displacement curve of the current environmental condition, the simulation average resistance and the actual average resistance are calculated respectively, and the average resistance error between the simulation average resistance and the actual average resistance is calculated.

[0116] Exemplarily, the simulation average resistance can be obtained by averaging all resistance values in the simulation total resistance-displacement curve, that is, , wherein, represents the simulation average resistance. Similarly, the calculation formula of the actual average resistance is . The absolute error between the simulation average resistance and the actual average resistance is calculated, that is, , wherein A represents the average resistance error. The average resistance error is used to measure whether the overall level of the simulation curve is consistent with the measured curve.

[0117] S414, the root mean square error, the peak error and the average resistance error are weighted and summed to obtain the total error. The sum of the weights corresponding to the root mean square error, the peak error and the average resistance error is 1.

[0118] Exemplarily, the calculation formula of the total error is , wherein E represents the total error, , , respectively, are the weights of the root mean square error, the peak error and the average resistance error, which can be set according to the importance of the error, and satisfy . For example, =0.5, = 0.3, = 0.2, RMSE has the greatest impact on the overall fitting, and the weight can be higher; the peak error has a large impact on durability and transient, and the weight can be moderate; the average resistance error has a moderate impact on the smooth experience of use, and the weight can be smaller.

[0119] S415, the total error of the four environmental conditions is weighted and averaged to obtain the simulation error.

[0120] Exemplarily, the total error of the four environmental conditions can be weighted and averaged to obtain the simulation error. For example, , wherein, represents the total error of the jth environmental condition, and the errors of the four environmental conditions can be weighted equally or different weights can be given to different environmental conditions according to actual conditions.

[0121] Through RMSE, peak error and average resistance error, the difference between simulation results and actual test data can be comprehensively evaluated to ensure the accuracy of the simulation model in multiple aspects. Compared with single error measurement method, multiple error measurement methods are more detailed and comprehensive, which can effectively avoid focusing on the error of a certain interval or stage, and provide more reliable simulation results.

[0122] S420, error comparison: compare the simulation error with the error threshold to determine whether the simulation error is less than the error threshold.

[0123] Exemplarily, for simulation in the fields of precision machinery, aerospace, etc., the error is required to be very small, and the error threshold can be set to between 1% and 2%; while for general consumer product design or conventional engineering application, the error tolerance is larger, and the error threshold can be set to between 5% and 10%. For the waterproof zipper of the bag, the error threshold can be set to 5% to 10%.

[0124] The simulation error can be compared with the error threshold. If the simulation error is less than the error threshold (i.e. ), it can be considered that the key parameters meet the actual situation, and the optimization process can be ended; if the simulation error is greater than or equal to the error threshold (i.e. ), the parameter optimization is continued.

[0125] S430, parameter optimization: when the simulation error is greater than or equal to the error threshold, based on the simulation error and the parameter set, each key parameter in the parameter set is optimized through an optimization algorithm to obtain an optimized parameter set. The parameter set includes the hyperelasticity parameters of the sealing strip at room temperature, the hyperelasticity parameters of the sealing strip at low temperature, the friction coefficient between the sealing strips in dry state, the friction coefficient between the sealing strips in wet state, the friction coefficient between the chain teeth in dry state, and the friction coefficient between the chain teeth in wet state.

[0126] For example, the hyperelasticity parameter of the sealing strip at room temperature ( , ), hyperelasticity parameters of sealing strips at low temperatures ( , ), coefficient of friction between sealing strips in dry state ( ), coefficient of friction between sealing strips in wet conditions ( ), the coefficient of friction between chain teeth in the dry state ( ), the coefficient of friction between chain teeth in wet conditions ( ) Construct as a parameter set For example, an initial set of parameters can be constructed from the example in step S230, i.e. .

[0127] Each parameter can be updated by calculating the gradient of the objective function with respect to each parameter in the parameter set using gradient descent. ,in, This represents the updated value of the k-th parameter. This represents the current value of the k-th parameter. This represents the gradient of the objective function with respect to the k-th parameter.

[0128] Optionally, S430, based on simulation errors and the parameter set, optimizes each key parameter in the parameter set using an optimization algorithm to obtain an optimized parameter set, including:

[0129] S431, based on simulation errors, parameter sets, and historical data, constructs a prediction mean function and a prediction variance function using Gaussian process regression. The historical data includes the simulation errors and parameter sets for each iteration.

[0130] It is understandable that, due to the extremely high cost of simulation (four simulations are required for one iteration), optimization cannot directly rely on brute-force search. Gaussian process regression (GPR) can be used to construct a regression model from the parameter set to the simulation error, thereby providing predictions for the optimization algorithm.

[0131] Gaussian process regression is a nonparametric method based on Bayesian theory used to learn latent functional relationships from limited training data.

[0132] For example, in the initial loop, there is only one set of parameters. and simulation error Gaussian process regression can still construct effective preliminary mean and variance functions. As more data is introduced, the prediction mean and variance functions gradually become more accurate.

[0133] The Gaussian process measures the similarity between two points in the parameter space by a kernel function, which can be a squared exponential kernel (RBF kernel) with the expression where is the signal variance, which controls the amplitude of the function output, is the length scale, which controls the smoothness of the function, is the Euclidean distance between the parameter set and . Since there is only one parameter set , the kernel value calculated by the squared exponential kernel is . A noise term can be added to calculate the inverse of the kernel value, i.e. where is the noise variance, which represents the noise level of the historical data, and I is the identity matrix with a size of 1x1.

[0134] Based on the squared exponential kernel and the parameter set , a covariance function vector can be constructed to represent the covariance (i.e., the kernel value) between the new parameter set and the parameter set , i.e. where represents the new parameter set (i.e., each parameter set generated during the optimization process).

[0135] The error between the transpose of the covariance function vector and the inverse of the kernel value and the simulation error can be calculated to obtain the preliminary prediction mean function, i.e. where represents the transpose of .

[0136] The preliminary prediction variance function can be constructed based on the transpose of the covariance function vector and the inverse of the kernel value, i.e. where represents the kernel value of the new parameter set with itself.

[0137] Exemplarily, at step S431, the prediction mean function and the prediction variance function are constructed by Gaussian process regression based on the simulation error, the parameter set, and the historical data, including:

[0138] S4311, a parameter matrix and an error vector are constructed based on the simulation error, the parameter set, and the historical data.

[0139] Exemplarily, it can be assumed that after N iterations, the simulation error and the parameter set have a total of N+1 groups, i.e. All parameter sets can be organized into a parameter matrix whose size is (N+1) x 8, and each row is an 8-dimensional parameter set, i.e. wherein, denotes the parameter set of the i-th iteration.

[0140] All the simulation errors are organized into an error vector F whose size is (N+1) x 1, and each element is the corresponding simulation error, i.e. wherein, denotes the simulation error of the i-th iteration.

[0141] S4312, based on the parameter matrix and the square exponential kernel, a kernel matrix is constructed, and an inverse matrix of the kernel matrix is calculated.

[0142] Exemplarily, according to the square exponential kernel function, the kernel function value between each two parameter sets in the parameter matrix (including each parameter set itself) can be calculated, and all the calculated kernel function values are constructed into a kernel matrix, i.e. wherein, the size of the kernel matrix K is (N+1) x (N+1), and the kernel matrix K is used to represent the similarity matrix between the historical data points.

[0143] The kernel matrix K can become close to singular due to numerical errors or too high similarity of data, so that the calculation of (the inverse of the kernel matrix K) becomes difficult or unstable. In order to facilitate the stability of matrix calculation and avoid singular matrix problem, a small noise term can be added to the kernel matrix K, and the inverse matrix of the kernel matrix K can be represented as wherein, I is an identity matrix with a size of (N+1) x (N+1).

[0144] In practical applications, the noise variance can be determined by maximum likelihood estimation (MLE), and the likelihood function can be represented as by solving , the noise variance can be obtained.

[0145] S4313, based on the parameter matrix and the square exponential kernel, a covariance function vector is constructed.

[0146] Exemplarily, the covariance function vector is used to calculate the kernel function value between the new parameter set and each parameter set in the parameter matrix. The covariance function vector can be represented as .

[0147] S4314, based on the covariance function vector, the inverse matrix of the kernel matrix and the error vector, a prediction mean function is constructed.

[0148] Exemplarily, the prediction mean function can be represented as wherein, denotes the transpose of the covariance function vector, The constant vector is calculated once. The predictive mean function can predict the new parameter set corresponding simulation error.

[0149] S4315, according to the covariance function vector and the inverse matrix of the kernel matrix, the predictive variance function is constructed.

[0150] Exemplarily, the predictive variance function can measure the uncertainty (i.e. confidence) of the corresponding simulation error of the new parameter set. The predictive variance function can be expressed as .

[0151] The above steps construct two predictive models from the parameter set to the simulation error through Gaussian process regression, which not only accurately predict the simulation error of the new parameter set, but also quantify the uncertainty of the prediction, helping the optimization algorithm to search for the optimal solution in the parameter space more efficiently, while avoiding excessive exploration of unnecessary areas, thereby improving the optimization accuracy and efficiency of the model.

[0152] S432, based on the predictive mean function and the predictive variance function, the acquisition function is constructed.

[0153] Exemplarily, the preliminary predictive variance function can be converted into a preliminary predictive standard deviation function According to the preliminary predictive mean function, the preliminary predictive standard deviation function and the simulation error , a preliminary variable function is constructed, that is , wherein represents a hyperparameter, which is used to balance the trade-off between exploration (exploring areas with large uncertainty) and exploitation (exploiting areas near the current optimal solution). It can be set to a small positive number (such as 0.01 or 0.1).

[0154] If the preliminary variable function is subject to a Gaussian distribution, the probability density function can be expressed as ; the cumulative distribution function can be expressed as , wherein represents the cumulative function of the standard normal distribution. The cumulative distribution function of the standard normal distribution can be expressed as .

[0155] The preliminary acquisition function can be expressed as .

[0156] Exemplarily, step S432, based on the predictive mean function and the predictive variance function, the acquisition function is constructed, including:

[0157] S4321, the minimum value is extracted from the simulation error and the simulation error in the historical data, and the optimal error is obtained. ​

[0158] Exemplarily, the optimal error is the minimum value among all simulation errors in N iterations, and the optimal error can be used to compare with the subsequent predicted mean and predicted standard deviation, thereby guiding the optimization process. The minimum simulation error can be extracted from all simulation errors obtained in N iterations, that is, , wherein, represents the optimal error.

[0159] S4322, convert the predicted variance function into a predicted standard deviation function, and construct a variable function according to the predicted mean function, the predicted standard deviation function and the optimal error.

[0160] Exemplarily, the predicted standard deviation function is the square root of the predicted variance function, and the predicted standard deviation function is , the predicted standard deviation function is used to measure the volatility of the prediction.

[0161] The constructed variable function (which can also be called an improvement amount) is .

[0162] S4323, construct a probability density function and a cumulative distribution function according to the variable function.

[0163] It can be understood that in Gaussian process regression, the probability density function (PDF) and the cumulative distribution function (CDF) can be used to quantify the potential improvement of the variable function.

[0164] Exemplarily, the variable function is subject to a Gaussian distribution, and the probability density function can be expressed as . The probability density function is used to calculate the probability distribution of the improvement amount of the new parameter set .

[0165] The cumulative distribution function can describe the probability that a random variable takes a value less than or equal to a certain specific value, and the cumulative distribution function can measure the improvement probability of the new parameter set with respect to the current optimal solution, thereby guiding the optimization algorithm to search more effectively. The cumulative distribution function can be expressed as .

[0166] S4324, construct an acquisition function according to the optimal error, the predicted mean function, the predicted standard deviation function, the probability density function and the cumulative distribution function.

[0167] It can be understood that in Bayesian optimization, the acquisition function is used to select the next evaluation point (i.e. the new parameter set). The goal of the acquisition function is to balance exploration and exploitation, and the expected improvement (EI) can be used as a specific implementation of the acquisition function. The expected improvement not only considers the predicted mean of the new parameter set, but also considers the difference between the predicted standard deviation and the current optimal solution.

[0168] Exemplarily, the acquisition function can be expressed as .

[0169] The acquisition function constructed by the above steps can effectively balance exploration and utilization, thereby guiding the optimization algorithm to efficiently search for the optimal solution and improving optimization efficiency and stability.

[0170] S433, based on the acquisition function, the optimal value of each key parameter in the parameter set is searched within the preset parameter boundary by an optimization algorithm, and an optimized parameter set is obtained. The preset parameter boundary includes the parameter range of the hyperelasticity parameter of the sealing strip, the parameter range of the friction coefficient between the sealing strips, and the parameter range of the friction coefficient between the chain teeth.

[0171] Exemplarily, considering material batch variation and temperature change, the parameter range of C10 and C01 can be set to 0.5 to 2 times (or adjusted according to actual conditions) of the test value (i.e. initial value). For example, C10 is 0.8 and C01 is 0.2 at room temperature, and the parameter range of C10 and C01 at room temperature can be set to , . For low temperature conditions, a wider range can also be considered according to the low temperature test data, because the material performance changes more at low temperature, such as , .

[0172] The typical friction coefficient range can be obtained from engineering manuals, material databases or published literature. For example, the reference value of the friction coefficient: the parameter range of rubber and rubber (such as sealing strip) under dry state condition is [0.5, 0.8], and the parameter range under wet state condition is [0.1, 0.3]; the parameter range of plastic and plastic (such as plastic zipper) under dry state condition is [0.15, 0.25], and the parameter range under wet state condition is [0.05, 0.15].

[0173] Since the method adopted in the initial value determination stage of the friction coefficient is a friction test, at least four friction tests can be performed on the sealing strip part and the sprocket part under dry and wet conditions, respectively, and the average value and standard deviation of the test data are calculated, and the parameter range can be set as the average value ± 3 times the standard deviation to ensure a confidence level of 99.7%. For example, for the sealing strip part, the friction coefficient test values under dry conditions are 0.35, 0.40, 0.38, and 0.42, so the average value is 0.39 and the standard deviation is 0.03, and the parameter range of the sealing strip friction coefficient under dry conditions is 0.39 ± 0.09, i.e. [0.30, 0.48], which is rounded to [0.30, 0.50]; the friction coefficient test values under wet conditions are 0.15, 0.18, 0.16, and 0.17, so the average value is 0.165 and the standard deviation is 0.012, and the parameter range of the sealing strip friction coefficient under wet conditions is 0.165 ± 0.036, i.e. [0.129, 0.201], which is rounded to [0.10, 0.25]. The same applies to the sprocket part, such as the parameter range of the sprocket friction coefficient under dry conditions is [0.15, 0.35], and the parameter range of the sprocket friction coefficient under wet conditions is [0.05, 0.25].

[0174] It can be understood that Bayesian optimization is a global optimization method based on acquisition functions, which guides the search by using the prediction results and uncertainties of the prediction mean function and the prediction standard deviation function to gradually find the optimal solution. The core idea of Bayesian optimization is to select the next evaluation point (i.e. a new parameter set) through the acquisition function, and update the prediction mean function and the prediction standard deviation function in each iteration.

[0175] For example, at the beginning of the optimization, there is already a set of parameter sets and simulation errors and a preliminary prediction mean function , a preliminary prediction standard deviation function . The acquisition value of each possible position within the preset parameter boundary can be calculated by the acquisition function , and the point with the maximum acquisition value is selected as the next evaluation point, and the new parameter set is the optimized parameter set.

[0176] For example, in step S433, based on the acquisition function, the optimal value of each key parameter in the parameter set is searched within the preset parameter boundary by an optimization algorithm to obtain an optimized parameter set, including:

[0177] S4331, based on the preset parameter boundary, a candidate point set is generated by a sampling method. The candidate points include the hyperelastic parameters of the sealing strip at room temperature, the hyperelastic parameters of the sealing strip at low temperature, the friction coefficient between the sealing strips under dry conditions, the friction coefficient between the sealing strips under wet conditions, the friction coefficient between the sprocket under dry conditions, and the friction coefficient between the sprocket under wet conditions.

[0178] Exemplarily, the sampling method includes Latin hypercube sampling (LHS), grid search, random sampling, and in this step, Latin hypercube sampling is selected. LHS can uniformly sample in the entire parameter space and is suitable for multi-dimensional parameter optimization problems.

[0179] For each key parameter (i = 1, 2, 3, 4, 5, 6, 7, 8, i.e. ), the parameter range is , and the parameter range can be divided into n sub-intervals. For example, the parameter range of is , which is divided into n equal sub-intervals: , where represents the boundary of the jth sub-interval after division.

[0180] In each sub-interval, a point can be randomly selected to generate a sampling point. For example, in the jth sub-interval of , a point is randomly selected, so that each sub-interval has a representative sampling point.

[0181] The above operation can be performed on the parameter range of each key parameter to generate n candidate point sets. Each candidate point is a vector composed of sampling points of each key parameter. The candidate point can be represented as .

[0182] Each candidate point in the candidate point set contains the value of the key parameter and can represent different regions of the entire parameter space (preset parameter boundary). The candidate point set provides multiple candidate points for the subsequent optimization process, and the candidate points will be used to evaluate the model to find the optimal solution.

[0183] S4332, based on the acquisition function, calculate the function value of each candidate point in the candidate point set, and sort the function value of each candidate point from high to low, and select the top N points as the initial point set.

[0184] Exemplarily, each candidate point in the candidate point set can be substituted into the acquisition function , and the function value of each candidate point can be calculated. The function values can be arranged in descending order, and the candidate points with the top N function values can be selected as the initial point set, where N can be 3-10 (depending on the dimension).

[0185] This step provides an initial exploration area for subsequent optimization, helping the optimization algorithm to focus on the parameter region with the largest potential improvement.

[0186] S4333, performing local optimization on each initial point in the initial point set by a local optimization algorithm to obtain a final point set.

[0187] It can be understood that the objective of the local optimization algorithm is to perform local search on each initial point and finely adjust each initial point to find a better solution.

[0188] In a high-dimensional parameter space, the uncertainty of the model (especially in the early stage of model training) can be very high. In this case, selecting the candidate point with the maximum function value in the candidate point set as the optimized parameter set can focus attention on the area with large prediction uncertainty. Further adjusting the parameters around the initial point through local optimization helps to reduce the prediction uncertainty and thus improve the reliability of the search result.

[0189] Local optimization can fine-tune the initial point by searching a small range, using existing information and local gradients to help the algorithm find a more accurate solution than the initial predicted optimal point, avoiding errors that can be caused by relying solely on the results of global sampling.

[0190] By way of example, the local optimization algorithm includes gradient descent, Powell method, BFG algorithm, and Nelder-Mead algorithm.

[0191] An initial point can be selected from the initial point set, the target function value of the initial point is calculated, and the value of the initial point is adjusted according to the selected local optimization algorithm (such as Nelder-Mead). The adjustment process is performed along the gradient of the target function or the quasi-Newton direction, and the specific manner depends on the optimization algorithm. It can be determined whether the optimization process converges, and if it converges, the optimization is stopped; otherwise, the local optimization process is continued. The convergence condition can be that the target function value is less than a preset threshold or that the maximum number of iterations is reached.

[0192] The above process is performed on each initial point in the initial point set to obtain a final point set.

[0193] Optionally, step S4333, performing local optimization on each initial point in the initial point set by a local optimization algorithm to obtain a final point set, includes:

[0194] For each initial point in the initial point set:

[0195] S43331, gradient calculation: calculating the current gradient based on the initial point and the acquisition function.

[0196] By way of example, the current gradient can be calculated approximately by numerical differentiation (such as finite difference) or analytically. The formula for calculating the current gradient is wherein, denotes the current gradient.

[0197] S43332, updating: based on the current gradient, the initial point is updated to obtain an updated initial point.

[0198] Exemplarily, the gradient descent method can be used to update the initial point according to the current gradient, and the calculation formula of the gradient descent method is wherein, denotes the updated initial point, denotes the current initial point.

[0199] S43333, range limitation: the updated initial point is range-limited according to the preset parameter boundary to obtain a limited initial point.

[0200] Exemplarily, since the update of the initial point can cause some key parameters to exceed the preset parameter boundary, the updated initial point can be range-limited to facilitate each key parameter to remain within the effective range.

[0201] For each key parameter , if the updated value exceeds the corresponding parameter range, it can be limited within the parameter range, i.e. .

[0202] S43334, iteration gradient calculation step, updating step, range limitation step until the convergence condition is reached to obtain the final point. The convergence condition is that the iteration number reaches the preset iteration number or the current gradient norm is less than the gradient threshold.

[0203] Exemplarily, steps S43331, S43332 and S43333 can be repeatedly executed for each initial point to determine whether the convergence condition is met. When one of the convergence conditions is met, the optimization process stops, and the final point corresponding to each initial point is obtained.

[0204] The convergence condition can be that the iteration number reaches the preset iteration number, and then the optimization process stops. For complex problems, a higher iteration number (such as 100, 200 times) can be set; for relatively simple or not too complex optimization problems, an iteration number of 20-50 times can be set.

[0205] The convergence condition can be that the current gradient norm is less than the gradient threshold, i.e. wherein, denotes the gradient threshold. The gradient threshold can be set to a small value, such as or If the computing resources are limited or do not require very high accuracy, a larger gradient threshold such as or .

[0206] Through multiple iterations, a more accurate solution can be found within a preset number of iterations or when the gradient change tends to be stable, thereby effectively improving the optimization accuracy and facilitating the obtaining of an optimized result that meets the preset parameter boundary. Through the above steps, global search and local adjustment can be balanced in a complex parameter space, thereby improving the overall optimization efficiency and stability of the result.

[0207] S4334, according to the acquisition function, calculate the function value of each final point in the final point set, and select the final point corresponding to the maximum value from the function value of each final point as the optimized parameter set.

[0208] Exemplarily, each final point in the final point set can be substituted into the acquisition function , that is, the function value of each final point can be calculated, all the function values of the final points are sorted from high to low, and the final point corresponding to the first function value (i.e., the maximum function value) is selected as the optimized parameter set.

[0209] This step is beneficial to the optimization algorithm to select the parameter set with the maximum potential improvement, and provides an optimal starting point for further optimization. Through the above steps, global search and local fine optimization can be efficiently balanced, the model accuracy can be improved and the convergence can be accelerated by gradually selecting the optimal solution, thereby obtaining the optimal parameter set.

[0210] S440, simulation test: based on the optimized parameter set, the simulation model is simulated and tested to obtain the simulation result.

[0211] Exemplarily, based on the optimized parameter set, the key parameters corresponding to each environmental condition are modified in the simulation model. Under the same test conditions (test time, travel), steps S310 and S320 are repeated to calculate the new total resistance-displacement curve under each environmental condition.

[0212] S450, iteration error calculation step, error comparison step, parameter optimization step, simulation test step until the simulation error is less than the error threshold or the number of iterations reaches the preset number of iterations, and the calibrated simulation model is obtained.

[0213] Exemplarily, steps S410, S420, S430, and S440 are repeatedly executed when or the preset maximum number of iterations (such as 100 iterations) is reached, the optimization process is ended, and the calibrated parameter set under each environmental condition is obtained.

[0214] By iteratively optimizing the key parameters of the simulation model, the error between the simulation result and the actual test data is gradually reduced, thereby improving the accuracy of the simulation model. And make the simulation model can accurately predict the performance of the waterproof zipper in various environmental conditions, provide reliable product design basis.

[0215] S500, based on the test time and the test stroke, simulating the calibrated simulation model to obtain the engagement resistance of the waterproof zipper of the luggage to be tested.

[0216] Exemplarily, after the calibration of the key parameters in the simulation model is completed, the final optimized key parameters can be used to re-simulate the simulation model. During the simulation test, the final optimized key parameters under different environmental conditions can be replaced to obtain the total resistance-displacement curve, the engagement resistance-displacement curve, and the sealing resistance-displacement curve under four environmental conditions.

[0217] It can be assumed that the test stroke of the waterproof zipper of the luggage is 200 mm, and the values are taken at 6 displacement points, i.e. 0 mm, 40 mm, 80 mm, 120 mm, 160 mm, and 200 mm. The total resistance-displacement curve in the actual test data of the waterproof zipper of the luggage and the total resistance-displacement curve, the engagement resistance-displacement curve, and the sealing resistance-displacement curve in the simulation result are shown in Figure 2 、 Figure 3 、 Figure 4 、 Figure 5 as shown in the table (in the form of a table).

[0218] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0219] The electronic device provided in the embodiments of the present application includes at least one processor, at least one memory, and a computer program stored in the at least one memory and executable on the at least one processor. When the processor executes the computer program, the electronic device implements the steps in any of the above electronic method embodiments.

[0220] Exemplarily, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. One or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0221] The electronic device can be a desktop computer, a notebook computer, a palm computer, or the like. The electronic device can include, but is not limited to, a processor and a memory. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic, discrete hardware components, or the like. The memory can be an internal storage unit of the electronic device, such as a hard disk or a memory of the electronic device, or can be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, or the like. Further, the memory can include both the internal storage unit and the external storage device. The memory is used to store an operating system, application programs, a boot loader, data, and other programs, such as program codes of computer programs. The memory can also be used to temporarily store data that has been output or will be output.

[0222] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0223] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit the same; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for determining the engagement resistance of a waterproof zipper for bags, characterized in that, include: Obtain actual test data and a 3D model of the waterproof zipper of the bag to be tested; wherein, the actual test data includes total resistance-displacement curves, test time, and test stroke under four environmental conditions, the four environmental conditions include normal temperature dry state, normal temperature humid state, low temperature dry state, and low temperature humid state, and the total resistance includes engagement resistance and sealing resistance; Based on the aforementioned 3D model, a simulation model of the waterproof zipper of the bag under test is constructed. The simulation model includes key parameters and contact pairs. The key parameters include the hyperelasticity parameter of the sealing strip, the friction coefficient between sealing strips, and the friction coefficient between chain teeth. Four environmental conditions correspond to four sets of key parameters. The contact pairs include meshing contact pairs and sealing contact pairs. The meshing contact pairs are used to simulate the mechanical meshing interaction between chain teeth and between chain teeth and the zipper pull. The sealing contact pairs are used to simulate the compression sealing interaction between sealing strips. Based on the test time and the test journey, the simulation model is subjected to simulation test to obtain simulation results; wherein, the simulation results include the total resistance-displacement curve, meshing resistance-displacement curve, and sealing resistance-displacement curve under the four environmental conditions; Based on the simulation results and the actual test data, the key parameters of the simulation model are iteratively optimized to obtain the calibrated simulation model. Based on the test time and the test stroke, the calibrated simulation model is subjected to simulation test to obtain the engagement resistance of the waterproof zipper of the bag under test.

2. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 1, characterized in that, The process of constructing a simulation model of the waterproof zipper of the bag under test based on the three-dimensional model includes: The three-dimensional model is geometrically cleaned and simplified to obtain the first model; The first model is meshed and its local mesh is refined to obtain the second model; Material properties are added to the second model, and contact pairs and boundary conditions of the second model are defined to obtain the simulation model.

3. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 1, characterized in that, The simulation test is performed on the simulation model based on the test time and the test schedule to obtain simulation results, including: Based on the test time and the test distance, construct a displacement function; For each environmental condition: Based on the displacement function and key parameters under the current environmental conditions, the simulation model is solved by a solver to obtain the total resistance-displacement curve, meshing resistance-displacement curve, and sealing resistance-displacement curve under the current environmental conditions.

4. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 1, characterized in that, The step of iteratively optimizing the key parameters of the simulation model based on the simulation results and the actual test data to obtain a calibrated simulation model includes: Error calculation: Based on the objective function, calculate the simulation error between the total resistance-displacement curve in the simulation results and the total resistance-displacement curve in the actual test data; Error comparison: The simulation error is compared with an error threshold to determine whether the simulation error is less than the error threshold; Parameter optimization: When the simulation error is greater than or equal to the error threshold, based on the simulation error and the parameter set, each key parameter in the parameter set is optimized using an optimization algorithm to obtain an optimized parameter set; wherein, the parameter set includes the hyperelasticity parameter of the sealing strip at room temperature, the hyperelasticity parameter of the sealing strip at low temperature, the friction coefficient between sealing strips in a dry state, the friction coefficient between sealing strips in a wet state, the friction coefficient between chain teeth in a dry state, and the friction coefficient between chain teeth in a wet state; Simulation test: Based on the optimized parameter set, the simulation model is simulated and tested to obtain the simulation results; The process involves iterating through the error calculation step, the error comparison step, the parameter optimization step, and the simulation testing step until the simulation error is less than the error threshold or the number of iterations reaches the preset number of iterations, thereby obtaining the calibrated simulation model.

5. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 4, characterized in that, The step of calculating the simulation error between the total resistance-displacement curve in the simulation results and the total resistance-displacement curve in the actual test data according to the objective function includes: For each environmental condition: The root mean square error is calculated based on each data point of the simulated total resistance-displacement curve and each data point of the measured total resistance-displacement curve under the current environmental conditions; wherein, the simulated total resistance-displacement curve is the total resistance-displacement curve in the simulation results, and the measured total resistance-displacement curve is the total resistance-displacement curve in the actual test data; Extract the simulated peak resistance and the actual peak resistance from the simulated total resistance-displacement curve and the measured total resistance-displacement curve under the current environmental conditions, respectively, and calculate the peak error between the simulated peak resistance and the actual peak resistance; Based on the simulated total resistance-displacement curve and the measured total resistance-displacement curve under the current environmental conditions, calculate the simulated average resistance and the actual average resistance, and calculate the average resistance error between the simulated average resistance and the actual average resistance. The total error is obtained by weighted summation of the root mean square error, the peak error, and the average resistance error; wherein the sum of the weights corresponding to the root mean square error, the peak error, and the average resistance error is 1. The simulation error is obtained by weighted averaging of the total errors under the four environmental conditions.

6. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 4, characterized in that, Based on the simulation error and parameter set, an optimization algorithm is used to optimize each key parameter in the parameter set to obtain an optimized parameter set, including: Based on the simulation error, the parameter set, and historical data, a prediction mean function and a prediction variance function are constructed using Gaussian process regression; wherein, the historical data includes the simulation error and parameter set for each iteration; Based on the predicted mean function and the predicted variance function, a data acquisition function is constructed; Based on the acquisition function, the optimal value of each key parameter in the parameter set is searched within the preset parameter boundary by an optimization algorithm to obtain the optimized parameter set; wherein, the preset parameter boundary includes the parameter range of the hyperelasticity parameter of the sealing strip, the parameter range of the friction coefficient between the sealing strips, and the parameter range of the friction coefficient between the chain teeth.

7. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 6, characterized in that, The step of constructing a prediction mean function and a prediction variance function through Gaussian process regression based on the simulation error, the parameter set, and historical data includes: Based on the simulation error, the parameter set, and historical data, construct the parameter matrix and the error vector; Based on the parameter matrix and the squared exponent kernel, construct the kernel matrix and calculate the inverse of the kernel matrix; Construct a covariance function vector based on the parameter matrix and the squared exponent kernel; The prediction mean function is constructed based on the covariance function vector, the inverse of the kernel matrix, and the error vector. The prediction variance function is constructed based on the covariance function vector and the inverse of the kernel matrix.

8. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 6, characterized in that, The step of constructing a data acquisition function based on the predicted mean function and the predicted variance function includes: The minimum value is extracted from the simulation error and the simulation error in the historical data to obtain the optimal error; The prediction variance function is converted into a prediction standard deviation function, and a variable function is constructed based on the prediction mean function, the prediction standard deviation function, and the optimal error; Based on the aforementioned variable function, construct the probability density function and cumulative distribution function; The acquisition function is constructed based on the optimal error, the predicted mean function, the predicted standard deviation function, the probability density function, and the cumulative distribution function.

9. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 6, characterized in that, Based on the acquisition function, an optimization algorithm is used to search for the optimal value of each key parameter in the parameter set within a preset parameter boundary to obtain an optimized parameter set, including: Based on the preset parameter boundaries, a set of candidate points is generated by a sampling method; wherein, the candidate points include the hyperelasticity parameter of the sealing strip at room temperature, the hyperelasticity parameter of the sealing strip at low temperature, the friction coefficient between the sealing strips in the dry state, the friction coefficient between the sealing strips in the wet state, the friction coefficient between the chain teeth in the dry state, and the friction coefficient between the chain teeth in the wet state. Based on the acquisition function, calculate the function value of each candidate point in the candidate point set, sort the function values ​​of each candidate point from high to low, and select the first N points as the initial point set; By using a local optimization algorithm, each initial point in the initial point set is locally optimized to obtain the final point set; Based on the acquisition function, the function value of each final point in the final point set is calculated, and the final point corresponding to the maximum value of each final point is selected as the optimized parameter set.

10. The method for determining the engagement resistance of a waterproof zipper for bags as described in claim 9, characterized in that, The step involves using a local optimization algorithm to perform local optimization on each initial point in the initial point set to obtain the final point set, including: For each initial point in the set of initial points: Gradient calculation: Based on the initial point and the acquisition function, calculate the current gradient; Update: Based on the current gradient, update the initial point to obtain the updated initial point; Range restriction: Based on the preset parameter boundary, the updated initial point is range-restricted to obtain a restricted initial point; The gradient calculation step, the update step, and the range restriction step are iterated until the convergence condition is met to obtain the final point; wherein, the convergence condition is that the number of iterations reaches a preset number of iterations or the current gradient norm is less than the gradient threshold.

Images