Instance-based unlocking and locking mechanism tolerance optimization method and system

Abstract

The invention provides an instance-based unlocking and locking mechanism tolerance optimization method and system. Belongs to the field of intelligent firearm manufacturing. According to the method, on the basis of ensuring that the assembly precision is qualified, the machining cost and the procedure capability index are comprehensively considered, and the tolerance of the unlocking and locking mechanism is designed and optimized. According to the method, on the basis of an unlocking and locking mechanism tolerance instance library, through analysis and calculation of contribution degrees, product processing cost and quality loss optimization targets are constructed, a target optimization model with process capability indexes and tolerance accumulation as constraint conditions is established, and the processing cost and the assembly quality loss are balanced. The design precision under precision manufacturing can be met, the production efficiency is improved, the part quality is improved, and the production cost is further reduced.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent manufacturing of firearms, and particularly relates to a method and system for optimizing the tolerances of an unlocking and locking mechanism based on cases. Background Art

[0002] Assembly accuracy is an important quality index for measuring the level of assembly technology. During the actual use of a product, in order to enable components to perform their working functions normally, on the basis of meeting the assemblability, it is necessary to control the relative position accuracy, relative fit accuracy, and relative motion accuracy between components. During the process stage, the assembly accuracy is mainly controlled by adjusting the assembly process; while during the design stage, it is mainly ensured by the machining accuracy of the component design of the assembly body. However, improving the design accuracy of parts will increase the manufacturing cost of the product and affect the economy of the manufacturing process flow. Therefore, how to reasonably design and allocate tolerances is an important means to ensure assembly accuracy. Summary of the Invention

[0003] To solve the above technical problems, the present invention proposes a method and system for optimizing the tolerances of an unlocking and locking mechanism based on cases.

[0004] The first aspect of the present invention discloses a method for optimizing the tolerances of an unlocking and locking mechanism based on cases, and the method includes:

[0005] Step S1: Traverse the tolerance knowledge case base according to the new design requirements, calculate the requirement similarity between the new design requirements and each design case, and sort according to the requirement similarity, and select K design cases with a similarity higher than the similarity threshold as similar design cases; wherein, the tolerance knowledge case base stores multiple design cases, and each design case includes key measurement dimensions, nominal dimensions of components, and tolerance information; the key measurement dimensions and the nominal dimensions of components belong to the design requirements;

[0006] Step S2: Calculate the assembly accuracy according to the tolerance information in each similar design case, and then judge whether the tolerance information in any one of the similar design cases meets the new design requirements according to the calculated assembly accuracy; wherein,

[0007] If the tolerance information in the K similar design cases does not meet the new design requirements, perform the following steps:

[0008] Based on the objective optimization model of machining cost and quality loss function, with the process capability index and tolerance accumulation as constraint conditions, optimize the tolerance information in any one of the similar design cases to obtain a design scheme that meets the new design requirements.

[0009] Optionally, in the step S1, the KNN algorithm is used to calculate the requirement similarity between the new design requirements and each design case.

[0010] Optionally, in the step S2, the calculation formula for the assembly accuracy is:

[0011]

[0012] where T mi is the calculated assembly accuracy, T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; λ i is the weighting coefficient of the i-th tolerance.

[0013] Optionally, in the step S2, the target optimization model is:

[0014]

[0015] where min represents minimization; C is the objective function; ω i is the weighting coefficient converted according to the analysis of the contribution degree of component features; T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the component feature tolerance, where c0, c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the component feature quality loss function, where A is the limit loss of the product within the tolerance regulations, and T is the design tolerance.

[0016] Optionally, in the step S2, the constraint conditions include:

[0017]

[0018] C pk min ≤ C pk ≤ C pk max

[0019]

[0020] T i ≤ T

[0021] where C Pmin is the lower limit of the process capability index; C Pmax is the upper limit of the process capability index; C Pi is the process capability index corresponding to the i-th tolerance; σ is the standard deviation of the process in a stable state; C Pkmin is the lower limit of the actual process capability index; C Pkmax is the upper limit of the actual process capability index; C Pk is the actual process capability index. Among them, USL and LSL are respectively the upper and lower limits of the tolerance control requirements, which are the key measurement dimensions in the new design requirements. μ is the mean of the size distribution statistically obtained during the actual machining process; λ i is the weighting coefficient of the i-th tolerance, and T 0i is the assembly accuracy of the new design requirements.

[0022] Optionally, the method further includes:

[0023] S3. Store the design scheme of the new design requirements as a design instance in the tolerance knowledge instance library.

[0024] The second aspect of the present invention discloses an instance-based tolerance optimization system for an opening and closing mechanism. The system includes:

[0025] A first processing module, configured to traverse the tolerance knowledge instance library according to the new design requirements, calculate the requirement similarity between the new design requirements and each design instance, and sort according to the requirement similarity, and screen out K design instances higher than the similarity threshold as similar design instances. Among them, the tolerance knowledge instance library stores multiple design instances, and each design instance includes key measurement dimensions, nominal part dimensions, and tolerance information; the key measurement dimensions and nominal part dimensions belong to the design requirements;

[0026] A second processing module, configured to calculate the assembly accuracy according to the tolerance information in each similar design instance, and then judge whether the tolerance information in any similar design instance meets the new design requirements according to the calculated assembly accuracy. Among them,

[0027] If the tolerance information in the K similar design instances does not meet the new design requirements, perform the following actions:

[0028] Based on the objective optimization model of the processing cost and quality loss function, with the process capability index and tolerance accumulation as constraint conditions, optimize the tolerance information in any similar design instance to obtain a design scheme that meets the new design requirements.

[0029] In summary, the solution proposed by the present invention has the following technical effects: On the basis of ensuring qualified assembly accuracy, the method comprehensively considers the processing cost and the process capability index, and designs and optimizes the tolerance of the opening and closing mechanism. Based on the tolerance instance library of the opening and closing mechanism, through the analysis and calculation of the contribution degree and sensitivity, the optimization objectives of the product processing cost and quality loss are constructed, and an objective optimization model with the process capability index and tolerance accumulation as constraint conditions is established, so that the processing cost and assembly quality loss reach a balance, aiming to meet the design accuracy under precision manufacturing, improve production efficiency, improve part quality, and thus further reduce production costs. Description of the Drawings

[0030] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0031] Figure 1 It is a flowchart of a tolerance optimization method for an opening and closing mechanism based on examples according to an embodiment of the present invention;

[0032] Figure 2 It is a flowchart of the KNN algorithm according to an embodiment of the present invention;

[0033] Figure 3 It is a flowchart of tolerance design based on the K-nearest neighbor algorithm according to an embodiment of the present invention;

[0034] Figure 4 It is a schematic diagram of a contribution degree analysis calculation model according to an embodiment of the present invention;

[0035] Figure 5 It is a structural diagram of a tolerance optimization system for an opening and closing mechanism based on examples according to an embodiment of the present invention;

[0036] Figure 6 It is a structural diagram of an electronic device according to an embodiment of the present invention. Specific embodiments

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, rather than all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0038] According to an embodiment of the present invention, in a first aspect, a tolerance optimization method for an opening and closing mechanism based on examples is provided. Please refer to Figure 1 , and the method includes:

[0039] Step S1: Traverse the tolerance knowledge instance library according to the new design requirements, calculate the requirement similarity between the new design requirements and each design instance, and sort according to the requirement similarity. Select the top K design instances with a similarity higher than the similarity threshold as similar design instances. Among them, the tolerance knowledge instance library stores multiple design instances, and each design instance includes key measurement dimensions, nominal part dimensions, and tolerance information. The key measurement dimensions and nominal part dimensions belong to the design requirements.

[0040] Optionally, in step S1, the KNN algorithm is used to calculate the requirement similarity between the new design requirements and each design instance. The K-Nearest Neighbor algorithm is also known as the KNN algorithm. Its working principle is that if a sample can find that most of the K most similar (i.e., the nearest neighbors in the feature space) samples in the feature space belong to a certain category, then this sample also belongs to this category. That is, calculate the distances between the sample to be measured and all samples in the sample set according to the distance function, and then find the K samples closest to the sample to be measured according to the distance size, and judge the category of the sample to be measured based on these K samples. The KNN algorithm process is as Figure 2 shown.

[0041] The KNN algorithm is the simplest machine learning algorithm, and its theory is quite rich. Its advantage is that there is no need for training when using the KNN algorithm. Users only need to input samples, and the system will automatically classify the samples. When solving the sample to be measured, the system will automatically perform calculation and analysis and give the K samples closest to the sample to be measured. Therefore, the KNN algorithm is quite convenient to use and is suitable for multi-classification problems. However, this algorithm also has certain limitations. The KNN algorithm has problems of sample imbalance and overfitting. When there are a large number of samples in the sample set, the KNN algorithm tends to include samples into the KNN algorithm.

[0042] Considering that in the actual product R & D process, a large part of the products are optimized designs based on existing products, rather than starting from scratch. This application optimizes the firearm design process through the KNN algorithm. When facing the design requirements of new products, first calculate the instance similarity between the instances selected from the tolerance knowledge instance library and the problem to be solved, and sort according to the instance similarity. Select the nearest K similar instances and determine the instance category attribute of the new requirements. Finally, output the product design scheme for designers' reference and save the generated new instances to the tolerance knowledge instance library. The process is as Figure 3 shown.

[0043] In the design of firearms, this application establishes an intelligent design method for the tolerances of typical firearm mechanisms by using the KNN algorithm. Based on the previous knowledge and experience of the tolerances of mechanism components, the KNN algorithm is used to classify and screen the instances in the instance library, so as to realize the tolerance allocation scheme for the mechanism components of new firearm products, effectively improve the R & D efficiency of products, greatly shorten the R & D cycle, and improve the reliability of products.

[0044] Step 1a: Instance expression model based on matter-element analysis method

[0045] The expression of instances in the tolerance knowledge database is the most basic and primary problem to be solved based on the KNN classification algorithm. The quality of instance expression directly affects the quality of instance screening by the KNN classification algorithm. In the intelligent tolerance design system, the design features, design parameters, and design requirements involved in the tolerance design of each typical mechanism component are not the same. Therefore, the expression of instances in the tolerance knowledge instance library is a primary problem to be solved. In order to better express the tolerance knowledge instances, this application uses the matter-element analysis method to express the tolerance instances.

[0046] The matter-element analysis method is usually used to solve some difficult problems that cannot be solved by traditional mathematical analysis methods. The matter-element model usually describes things with an ordered triple. Let N be the object of study, C be the characteristics of the object of study, and X be the specific quantitative value corresponding to the characteristic C of the object of study i of the object of study, and the object of study, object characteristics, and quantitative value are called the three elements of the matter-element.

[0047] Suppose the instance object N m has n characteristics C = [c1, c2,..., c n , and the corresponding quantitative values are X = [x1, x2,..., x n . Then the matter-element model of the instance is expressed as:

[0048]

[0049] Among them, the instance matter-element model R is also called the n-dimensional matter-element.

[0050] Step 1b: Calculation of instance similarity

[0051] In the KNN algorithm, the concept of similarity is introduced to describe the distance between two instances. Usually, the Euclidean distance, cosine distance, and city block distance, etc. are used. In the KNN algorithm system, the traditional nearest neighbor classifier assigns the same weight to each attribute in the sample for similarity calculation. However, in some cases, when the feature is weakly or not related to the classification, it will mislead the classification result to have a large deviation.

[0052] Therefore, in order to better describe the similarity between instances, this application further calculates the similarity based on the instance expression model of the matter-element method. First, after determining the research object, then determine the functional feature elements of the instance and the target feature elements of the problem to be solved. Furthermore, allocate weights according to the importance of the feature units. Finally, solve the similarity value between this instance and the target problem to be solved.

[0053] Suppose the set of functional feature elements in a certain matter-element model contains n feature elements, then the instance functional feature set M in the instance library f and the target demand functional feature set M g of the matter-element models are respectively expressed as:

[0054] M f = <(C f , X f )>

[0055] = <(c f1 , c f2 , …, c fn ; x f1 , x f2 , …, x fn )>

[0056] M g = <(C g , X g )>

[0057] = <(c g1 , c g2 , …, c gn ; x g1 , x g2 , …, x gn )>

[0058] Therefore, this application establishes a corresponding target problem attribute system according to the actual demand problem M g , and assigns corresponding weights to each feature target element in the system according to its importance, so as to better describe the importance of the feature elements. Then the weight distribution relationship of each feature element in the target problem M g is expressed as:

[0059] W = (w1, w2, …, w n )

[0060] In the formula: 0 ≤ w k ≤ 1, and The weights can be solved by the analytic hierarchy process, which will not be elaborated here.

[0061] Suppose the similarity between the k-th functional feature element of a certain instance and the target feature is represented by es k , and es k∈[0, 1]. Among them, es k 's value is directly proportional to the similarity of the corresponding features of the instance. When es k takes the value of 1, it means that this instance completely matches the features of the target problem; on the contrary, when the features of the two do not match at all, es k has a value of 0; when the features of the two partially match, es k has a value between 0 and 1. Then the similarity degree between this instance and the target problem can be expressed as:

[0062]

[0063] In the formula: w k is the weight of the k-th attribute, n is the number of features of the current retrieved instance.

[0064] The similarity es k of the k-th feature element of the instance needs to be determined according to the distance between the value x fk of the functional feature element of the instance and the value x gk of the feature element of the target i. In the expression process of the instance, the values of the instance feature elements are mainly divided into two cases. One is a specific value, such as the tolerance precision level, the nominal size of the part, etc.; the other case is an interval, such as the value range of the tolerance value. Then the similarity calculation formula between the retrieved instance and the target instance needs to be analyzed and considered for different cases. As shown in Table 1 below, the different combination cases of the target feature value x gk and the retrieved instance feature value x fk are given in detail.

[0065] Table 1: Combination of Feature Element Value Types

[0066] Serial number <![CDATA[x gk > <![CDATA[x fk > 1 Determine the value Determine the value 2 Determine the value Interval 3 Interval Determine the value 4 Interval Interval

[0067] (1) When both x gk and x fk are determined values, the similarity es k between the feature elements is expressed as:

[0068]

[0069] In the formula: ρ represents the distance between x fk and x gk ; x fk is the feature value of the feature element k of a certain instance; x gk is the feature value of the feature element k in the target problem; and are the maximum and minimum values of the feature element k in the instance library.

[0070] If the attribute of the k-th feature element is a qualitative index, the qualitative index is converted into a quantitative index by manual assignment. Suppose the qualitative index of the part size is {poor, good, medium, excellent}, then the corresponding quantitative values are {0.25, 0.5, 0.75, 1}.

[0071] (2) When v fk is a definite value and v gk is an interval value, the instance similarity es k can be expressed by the extension distance calculation method:

[0072]

[0073] In the formula: ρ(x fk , x0, X gk ) is the distance between intervals; X fk (x f1 , x f2 ), X gk (x g1 , x g2 ) are the interval values of the instance function feature element and the target feature element respectively; x0 is the optimal value of the design interval; and are the maximum and minimum values of the k-th feature element in all instances.

[0074] To determine the optimal value x0, generally the method of taking the midpoint is adopted, that is Then the above formula can be changed to:

[0075]

[0076] For the case where it is difficult to solve the optimal value v0, it can be determined according to expert advice or previous design experience values, and then the value of the instance similarity es k is taken.

[0077] (3) When x fk is an interval value and x gk is a definite value, this situation accounts for a relatively small proportion in the actual application process, and the instance similarity is the same as the calculation method in (2).

[0078] (4) When x fk and x gk are both interval values, es k can be obtained by the following extension distance calculation method:

[0079]

[0080] In the formula: ρ(X fk , X gk ) is the distance between intervals; Xfk (x f1 ,x f2 )、X gk (x g1 ,x g2 ) are the interval values of the instance functional feature elements and the target feature elements respectively. and The maximum and minimum values of the k-th feature element in all instances.

[0081] In the process of instance reasoning, the above formula shows that the similarity between the instance and the target problem to be solved is directly proportional to Sim. The larger the calculated value of Sim, the higher the similarity between the instance and the problem to be solved, and the closer it is to the solution of the target problem.

[0082] Step 1c: Software implementation of the tolerance intelligent design module

[0083] The tolerance intelligent reasoning module of the typical firearm mechanism can greatly simplify the efficiency of product design and shorten the R & D cycle. Taking the locking mechanism as an example, its development process is as follows:

[0084] (1) Create an Access locking mechanism tolerance knowledge instance library

[0085] The locking mechanism tolerance database needs to clarify the type of knowledge stored in the database and the type of the database. This application uses an Access database with the file name ZhiNengGongCha.mdb. Subsequently, data tables of typical mechanisms such as the locking mechanism, shell ejection mechanism, and firing mechanism are stored in this database, and the field names included in the table respectively correspond to the input parameter string names of each mechanism. Through a large number of learning and research, the tolerance design situation and relevant expert experience in the tolerance analysis process of typical firearm mechanisms are analyzed. According to the firearm design manual and tolerance design criteria, the designed mechanism tolerance knowledge instance library contains the following types of characteristic information:

[0086] (a) Key measurement dimensions of technical requirements. Mainly used to ensure the feasibility of the mechanism and guarantee the design reliability of the mechanism.

[0087] (b) Nominal dimensions of mechanism components. Mainly includes certain dimensions in the mechanism components that affect the key measurement dimensions.

[0088] (c) Tolerance information. Information such as the upper and lower deviations and precision grades of the nominal dimensions of components.

[0089] Taking the intelligent design of the locking mechanism tolerance as an example, the locking clearance is the key measurement dimension for technical requirements. H1, H2, H3, D1, etc. are some nominal dimensions in the locking mechanism. H1ES, H2EI, and the accuracy grade of H1 are the upper and lower deviation and accuracy grade parameters of the nominal dimensions. According to the same method, the tolerance information of other nominal dimensions is sorted out and stored in this database to provide a retrieval source for subsequent intelligent tolerance design retrieval.

[0090] (2) Create a user interface

[0091] The inference interface based on the KNN algorithm is edited based on the.xml file. Enter the instance features of the target design in the interface, including key measurement dimensions, nominal dimensions, key information such as input and output, etc. The corresponding input in the module requires the feature requirements for retrieving instances. The information that needs to be provided by the designer is the basic parameters for incorporating the design requirements, mainly including the upper and lower limits of the locking clearance and the corresponding values of each nominal dimension. The output parameters are automatically generated by the module, including the upper and lower deviations and accuracy grades of each dimension.

[0092] (3) Create a KNN algorithm inference component

[0093] Since the retrieval of each mechanism in the firearm is carried out independently, and relatively speaking, the number of databases in each module is small, the KNN algorithm is adopted in this application for tolerance instance retrieval. And during the development of the software module, the meta-language XML used for information storage and transmission on the Internet is adopted for development and saved as a.config file. The XML language has absolute advantages in describing knowledge and the content of intelligent design information. Firstly, the form of knowledge description in the XML language is simple and easy to understand; secondly, it is more flexible and free in expressing tolerance knowledge, and can well represent the tolerance knowledge content of complex data; in addition, the XML language has high scalability in the content of tolerance knowledge, which is very convenient for expanding and improving the tolerance knowledge instance library; finally, the XML language has good openness and interoperability in the aspect of knowledge exchange, strong independence, can exist independently without relying on software programs, and can be recognized by multiple systems for knowledge sharing and realization.

[0094] (4) Run the inference component

[0095] After creating the inference component, run the inference component. The system will retrieve in the database according to the nearest neighbor retrieval algorithm, screen out 3 instances that are more in line with the current retrieval features, and pop them up in the form of a window dialog box for the designer to screen.

[0096] Among them, the instances output in the retrieval window are three groups of instances similar to the current problem retrieved from the database based on the KNN algorithm, and the order is sorted in descending order according to the similarity values of the strengths.

[0097] Case-based reasoning uses the K-nearest neighbor retrieval method to solve new problems. For features with the same numerical value, select the instance that is numerically closest; for string-type features, the similarity is judged by whether the strings are the same.

[0098] Step S2: Calculate the assembly accuracy based on the tolerance information in each similar design instance, and then judge whether the tolerance information in any similar design instance meets the new design requirements according to the calculated assembly accuracy; among them,

[0099] If the tolerance information in the K similar design instances does not meet the new design requirements, perform the following steps:

[0100] Based on the objective optimization model of machining cost and quality loss function, with the process capability index and tolerance accumulation as constraints, optimize the tolerance information in any similar design instance to obtain a design scheme that meets the new design requirements.

[0101] In this step, establish the 3D models of each component in the 3D modeling software, create the assembly sequence of the components with reference to the actual assembly process, define the relevant tolerance information according to the design requirements, set the key control dimensions as the analysis objectives and select appropriate simulation parameters for assembly simulation. Through the size distribution diagram, contribution factor analysis report and influence factor analysis report obtained from the simulation, comprehensively evaluate whether the control objectives meet the predetermined design assembly accuracy requirements. If the analysis results do not meet the assembly accuracy requirements or there is room for optimization, the designer can reallocate the tolerances according to the tolerance sensitivity analysis report on the premise of meeting the machining cost and quality requirements. Repeat the above process until the most optimized design scheme at this stage is obtained.

[0102] Optionally, in the step S2, the assembly accuracy calculation formula is:

[0103]

[0104] Among them, T mi is the calculated assembly accuracy, T i is the i-th tolerance value in the tolerance information, n is the number of tolerances in the tolerance information; λ i is the weighting coefficient of the i-th tolerance.

[0105] Optionally, in the step S2, the objective optimization model is:

[0106]

[0107] Among them, min represents minimization; C is the objective function; ω i is the weighting coefficient converted according to the component feature contribution degree analysis; T iis the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the part feature tolerance, where c0, c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the part feature quality loss function, where A is the limit loss when the product is within the tolerance, and T is the design tolerance.

[0108] Optionally, in the step S2, the constraint conditions include:

[0109]

[0110] T i ≤T

[0111] where C Pmin is the lower limit of the process capability index; C Pmax is the upper limit of the process capability index; C Pi is the process capability index corresponding to the i-th tolerance; σ is the standard deviation of the process in a stable state; C Pk min is the lower limit of the actual process capability index; C Pk max is the upper limit of the actual process capability index; C Pk is the actual process capability index, where USL and LSL are the upper and lower limits of the tolerance control requirements, which are the key measurement dimensions in the new design requirements, μ is the mean of the size distribution statistically obtained in the actual machining process; λ i is the weighting coefficient of the i-th tolerance, T 0i is the assembly accuracy of the new design requirements.

[0112] The quality loss function is the quality loss caused by the deviation of the product feature value from the target value. The quality loss increases with the square of the deviation. It is a loss function used to measure the quality of the model during the prediction process. It measures the accuracy of the model. The quality loss function consists of quality indicators, which can be data accuracy, prediction accuracy, production efficiency, etc. The basic principle of the quality loss function is: the gap between the actual prediction result and the model expected result, and this gap should be as small as possible to improve the efficiency and accuracy of the model prediction.

[0113] Contribution analysis is a method to quantitatively describe the influence of the input variables of a model on the change of output variables. After analyzing the three-dimensional assembly model of the locking mechanism, we can output the analysis results based on the HLM contribution analysis, including the contribution of the geometric features of each influencing factor to the locking clearance. Since the contribution analysis method is described based on a linear relationship, we approximately expand the Taylor formula to the first-order linear relationship expression, so that the HLM contribution can be calculated. The relationship between input and output quantitatively described in the contribution analysis model is as Figure 4 shown, and the general function expression f is used to describe the relationship between the output variable m and the input variable c i as follows:

[0114] m = f(c i ), i = 1, 2, …, t

[0115] Substitute the formula into the Taylor formula:

[0116]

[0117] In the formula, combined with the three-dimensional tolerance analysis model, m0 is the basic dimension of the characteristic tolerance; Δc i = c i - c i0 represents the change in the target dimension tolerance corresponding to a single tolerance change. Since the HLM contribution analysis method is an analysis and calculation of a linear scale, the Taylor formula can be approximately expanded to the first-order term, and the obtained linear function relationship is:

[0118]

[0119] When analyzing the contribution of one influencing factor, other influencing factors need to remain unchanged. To make the contribution analysis results credible, the following two assumptions must be met: 1) All influencing factors conform to the principle of statistical independence, that is, they have no correlation; 2) To simplify the model and calculation process, it is assumed that the distribution laws of the influencing factors C i all follow a normal distribution; under these two assumptions, we can further obtain the linear function relationship between the variance of the target dimension and the variances of the influencing factors:

[0120]

[0121] Then the calculation formula for the HLM contribution is:

[0122]

[0123] When calculating the first derivative , the contribution index can be used to replace it for calculation. According to the statistical simulation data of the target influencing factor, Ci The approximate substitution value. In the simulation software, set the tolerances of other influencing factors to 0, and only measure the change in the target size caused by the tolerance of one influencing factor. Repeat the above process and perform the same operation for each influencing factor. The corresponding maximum measured value can be obtained in the distribution pattern diagram of the target size. and the minimum measured value The value of C can be obtained through calculation i The variation of the target analyzed under the individual influence is: The HLM sensitivity index can be expressed as:

[0124]

[0125] is the sum of the variations under the individual actions of all influencing factors. Then, the above HLM contribution index can replace the first-order partial derivative to obtain:

[0126]

[0127] Furthermore, the formula for the final contribution can be obtained as:

[0128]

[0129] According to the contribution analysis report obtained from the research on the assembly accuracy modeling of key components, the information in Table 2 can be summarized:

[0130] Table 2: Simulation analysis results of HLM contribution

[0131]

[0132] Considering the actual situation, the influencing factors with a contribution less than 1% are omitted in the list of the HLM contribution decomposition results. After clarifying the percentage values of the HLM contributions of each characteristic plane, a preliminary optimization plan prototype can be made: 1) According to the contribution analysis results, considering that the contributions of the cartridge base plane and the nose plane of the gun body each account for 45.97%, which is much greater than the contributions of the mating surface at the tail end of the barrel body (2.87%) and the cartridge base plane of the cartridge case (1.84%) and the contribution of the mating surface at the tail end of the breechblock body (1.84%), the tolerances of the cartridge base plane and the nose plane of the gun body with greater influence can be directly optimized; on the other hand, considering that the effective tolerances of the cartridge base plane and the nose plane of the gun body are four to five times larger than those of the other three characteristic planes, and the control of the precision is relatively easy to combine with the actual processing technology. 2) Without considering the difficulty of processing, the weight coefficients can also be calculated according to the relative magnitudes of the contributions of each characteristic plane for target optimization, and the tolerances of the target size can be adjusted more scientifically to meet the design and use requirements. This plan is more suitable for target optimization where the contribution ratios of each influencing factor do not differ greatly.

[0133] When conducting tolerance design and optimization, the actual processing capacity level of the factory should be fully considered to make the tolerance distribution more economical and reasonable. Taking the process capability index as a constraint condition to achieve the tolerance optimization of products.

[0134] The process capability index, also known as the process capability ratio, describes the actual processing capacity of a process under continuous and stable production conditions. The size of the process capability has a great impact on the final product quality and usage reliability, and essentially reflects the level of the process yield. In the actual production process, when the process capability is stable, the dimensions of the machined parts basically conform to the normal distribution. According to the normal distribution diagram, the qualified rate of parts within 6σ can reach 99.73%. Therefore, we use the size of the 6σ value as the basis to reflect the size of the process capability value. When designing the part tolerance, it is generally required that the width of the size tolerance zone of the finished part (the tolerance range within 6σ) is less than or equal to the width of the design tolerance range, which is expressed by the formula:

[0135] PC = 6σ ≤ T

[0136] In the formula: PC (Product Specification) here refers to the technical requirements of the product tolerance constraint, σ is the standard deviation of the process in a stable state, and T is the total design tolerance.

[0137] To comprehensively evaluate the level of the process in production, the process capability index C p is used to reflect the size of the process capability:

[0138]

[0139] When the mean value of the product quality distribution characteristics does not coincide with the central value of the design tolerance, the allowances of the manufacturing tolerance with respect to the upper and lower limits of the design tolerance will be different, which may seriously affect the qualified rate of the product in severe cases. Therefore, it is necessary to correct the process capability value number C p The corrected actual process capability index is C pk :

[0140]

[0141] where USL and LSL are the upper and lower limits of the tolerance control requirements respectively, and μ is the mean value of the size distribution statistically in the actual processing process. Table 3 shows the process capability levels, and Table 4 shows the actual process capability levels.

[0142] Table 3: Process Capability Levels

[0143]

[0144] Table 4: Actual Process Capability Levels

[0145]

[0146] Through the statistical analysis of a large number of actual samples, to ensure the qualified rate of product processing and the subsequent assembly quality, the process capability should meet the minimum requirement C p > 1. C p > 1 does not guarantee that there will be no defective products, but it will control the defective rate within an acceptable range, C p The larger the value, the smaller the defective rate. Specifically, C p The adjustment and selection of the value are mainly related to the economic ability and the machinability ability.

[0147] The actual process capability index C pk is a measure value that describes the deviation degree between the size distribution center of the actual workpiece and the distribution center of the tolerance design standard on the basis of considering the process capability index C p . Ideally, it is hoped that the actual distribution center of the workpiece coincides with the central value of the design tolerance (the mean center of the tolerance normal distribution), so as to better ensure the qualified rate of parts and reduce the processing cost to a certain extent. Generally, the actual process capability index C pk complies with 1.67 > C pk ≥ 1.00, and the process capability is considered acceptable; C pk > 1.67, it is considered that the actual process capability is too high. An overly high actual process capability will greatly increase the manufacturing cost, and it is not necessary to reach this level for the production of ordinary parts; C pk ≤ 1, it is considered that the actual process capability is poor and cannot meet the control of the defective rate in the production process.

[0148] Processing cost function model

[0149] The total cost required in the process of part processing is called the processing cost. There is an irreconcilable contradiction between the processing cost and the manufacturing accuracy, and the manufacturing accuracy is mainly determined by the design accuracy. Therefore, the design accuracy is the most important factor affecting the processing cost. Selecting a strict tolerance grade during design can indeed better ensure the assemblability and functionality of parts, but the resulting cost problem cannot be ignored. On the basis of the research and analysis of the processing cost of typical firearm parts, different types of cost-tolerance model mathematical function relationships applicable to the tolerance design of typical firearm parts are established. This step intends to adopt a linear and exponential composite cost-tolerance model to express the tolerance cost functions of the plane features, outer circle features, inner hole features, runout tolerances, etc. of typical parts. As follows:

[0150]

[0151] Among them, c0, c1, c2, c3 are known parameters related to the tolerance.

[0152] The characteristic tolerance cost functions of typical parts include:

[0153] (1) The tolerance cost function for planar features is:

[0154]

[0155] (2) The tolerance cost function for external cylindrical features is:

[0156]

[0157] (3) The tolerance cost function for internal hole features is:

[0158]

[0159] (4) The runout tolerance cost function is:

[0160] C(T) = 0.0373e -3.08T

[0161] Step 4b: Quality loss cost function model

[0162] The cost model is based on the quality loss function. During the processing of products, affected by human factors and non - human random factors, the quality of products always fluctuates slightly above and below its target value. The amplitude of the fluctuation is mainly determined by the processing accuracy. Quality loss refers to the loss caused to users by quality fluctuations. Through the analysis of a large number of sample data, the size of quality loss can be approximately measured by a function.

[0163] If the quality characteristic value of the product is set as y and the expected value is m. When the quality characteristic value exactly satisfies y = m, it indicates that the product quality has no fluctuation and exactly meets the design expectation; while when y ≠ m, there is a certain gap between the product quality and the expected value. We use |y - m| to describe the difference between the two, and the size of the difference reflects the size of the quality loss. We expand the quality loss function L(y) at the expected value m according to the Taylor series:

[0164]

[0165] According to the description of the quality characteristics, when y = m, the product has no quality loss, corresponding to the quality loss function L(y) = 0; the quality loss takes the minimum value. According to the knowledge of advanced mathematics, at m, L′(m) = 0. Substituting it into the Taylor series expansion and omitting the higher - order terms of the quadratic term, we get the approximate value:

[0166] (1) Quality loss function for nominal - the - best characteristics:

[0167] L(y) = k(y - m) 2

[0168] In the formula, k is the mass loss coefficient. According to the design tolerance T of the product, it is stipulated that when |y - m| ≤ T, the product quality loss is within the tolerance range and the product is qualified; when |y - m| > T, the product quality loss exceeds the tolerance specified range and the product is unqualified. If the limit loss of the product within the tolerance is A, then A = kT 2 Furthermore, the mass loss coefficient can be obtained as k = A / T 2 .

[0169] (2) The quality loss function for the smaller-the-better characteristic is:

[0170] L(y) = ky 2

[0171] It is stipulated that when the quality characteristic value v satisfies 0 < v ≤ T, the product meets the requirements.

[0172] (3) The quality loss function for the larger-the-better characteristic is:

[0173] L(y) = k / y 2

[0174] It is stipulated that when the quality characteristic value v satisfies y ≥ T, the product meets the requirements.

[0175] Regarding the tolerance design, when the quality loss amount (y - m) is equal to the tolerance value T, the characteristic quality loss function can be transformed into:

[0176]

[0177] Then the total quality loss cost caused by the tolerance factor is:

[0178]

[0179] That is:

[0180]

[0181] In the formula, n is the number of tolerances in the product.

[0182] Previously, in the tolerance optimization stage, the tolerance allocation was generally carried out with the control of processing cost as the optimization goal, ignoring the impact caused by quality loss. This application establishes a hybrid optimization model of processing cost - quality cost loss, and converts the contribution degree analysis results into weighted coefficients and substitutes them into the optimization model for calculation. Considering the influence of tolerance factors on processing cost and quality loss comprehensively, with the premise of meeting the design tolerance requirements, the cost optimization is carried out with the minimum of processing cost and quality loss as the optimization goal. According to the analysis of the processing cost function model and quality loss cost model above, the total optimization objective function can be obtained as:

[0183]

[0184] Among them, min represents minimization; C is the objective function; ω i is the weighted coefficient converted according to the analysis of the contribution degree of component features; T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the component feature tolerance, where c0, c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the quality loss function of the component feature, where A is the limit loss of the product within the tolerance regulation, and T is the design tolerance.

[0185] When performing tolerance optimization, it is necessary to establish constraint conditions. Generally, factors such as the constraint relationship of the process capability index, the tolerance accumulation constraint relationship, and the value of the standard tolerance grade need to be considered. The process capability index and the assembly accuracy between mating parts with assembly accuracy requirements are both controlled within a certain range.

[0186] (1) Process capability index constraint conditions

[0187] To comprehensively evaluate the level of the process in production, the process capability index C p is used to reflect the size of the process capability. Different processing difficulties of the process have different requirements for the process capability index. Generally speaking, considering the processing cost, the process capability index of difficult-to-machine parts is greater than that of easy-to-machine parts. Therefore, when selecting a reasonable C p value, the actual processing ability and cost factors need to be comprehensively considered. The constraint conditions between the process capability index C p and the actual process capability index C pk are:

[0188]

[0189] σ is the standard deviation of the process in a stable state, T is the total design tolerance,

[0190] C pkmin ≤C pk ≤C pkmax

[0191]

[0192] where USL and LSL are the upper and lower limits of the tolerance control requirements respectively, and μ is the mean value of the size distribution statistically in the actual processing process.

[0193]

[0194] T0i is the assembly accuracy required by the design; λ i is the weighting coefficient;

[0195] T i ≤T

[0196] Generally, the assembly accuracy requirements are considered according to the actual processing situation. When 1 < C p < 1.33, the normal production of the product can be guaranteed. When C p > 1.67, the excessive process capability will cause great cost pressure. Therefore, when designing the tolerance, various factors of the parts are comprehensively analyzed and a suitable value is selected between 1 and 1.67. The value range of C pk is the same as the value range of C p .

[0197] (2) Tolerance accumulation constraint condition

[0198] During the assembly process, the fit between parts will transfer the machining errors of each part itself, and the errors will become larger and larger with the increase of the transfer times. This is the source of tolerance accumulation. The cumulative error of the final formed assembly clearance is the largest. In order to ensure the assembly accuracy, it is necessary to control the cumulative error of the dimensions. The constraint relationship is expressed as follows:

[0199]

[0200] In the formula, T 0i is the assembly accuracy required by the design; λ i is the weighting coefficient.

[0201] On the basis of ensuring the qualification rate of the assembly accuracy, the machining cost and the process capability index are comprehensively considered to optimize the tolerance of the automatic machine locking mechanism. After obtaining the results of the three-dimensional assembly accuracy simulation analysis, compare the requirements in the design stage and analyze whether the key control dimensions meet the design indicators. For the unqualified dimensions, convert the contribution analysis results into the weighting coefficient ω i , substitute it into the established hybrid model based on machining cost-quality loss for tolerance optimization, and convert the optimization results into standard tolerances and re-verify the optimization results after obtaining the optimization results.

[0202] The most ideal situation of the multi-objective optimization problem is that all component objectives reach the optimal solution at the same time. However, for tolerance optimization, due to the complexity of the part machining process and the functional requirements of the parts themselves, it is almost impossible for the tolerances to reach the optimal values at the same time. In this paper, a hybrid model with machining cost and quality loss as the optimization objectives is established, combined with the contribution analysis results and converted into the weighting coefficient, and the double-objective optimization is uniformly converted into a single-objective optimization to solve.

[0203] The three-dimensional tolerance model of the locking mechanism has been analyzed above. Now, taking the locking clearance as an example, the detailed process of tolerance optimization will be described.

[0204] According to the structural model of the parts and the actual machining situation, the following tolerance cost function can be established:

[0205]

[0206] The weighting factors are:

[0207]

[0208] Process capability index C P and the actual process capability index C Pk constraint conditions:

[0209]

[0210] 1 ≤ C pk = Min[USL - μ / 3σ, LSL - μ / 3σ] ≤ 1.67

[0211]

[0212] Single feature tolerance constraint conditions:

[0213]

[0214] According to the established tolerance optimization model, the following optimization results can be obtained:

[0215] Table 5: Optimization analysis results

[0216]

[0217] According to the optimization results, the converted standard tolerance is:

[0218] Table 6: Standard tolerance conversion results

[0219]

[0220] Substituting the optimized tolerance values into the assembly accuracy analysis model established in the three-dimensional analysis software, it can be obtained that the lower limit of the size distribution of the optimized locking clearance within ±3σ is 0.005 mm, and the upper limit is 0.129 mm, both of which are within the range of the assembly accuracy requirements, meeting the process capability index C p = 1.61, with good process capability, C pk = 1.33, with good actual process capability. The design indicators of the entire locking clearance basically meet the requirements. After optimization in the design stage, the tolerance distribution can be further adjusted according to the actual situation in the process stage.

[0221] In this application, in combination with the example of the latching clearance, a multi-objective optimization model of tolerance is constructed with the processing cost and quality loss as the optimization objectives and the process (or process) capability index and tolerance accumulation as the constraint conditions. The contribution degree analysis result of the latching clearance is converted into a weight coefficient and brought into the final optimization solution process, making the optimization model more reasonable and perfect. Finally, the optimization analysis result of the latching clearance size is controlled within the design requirement range, and the optimization of each key control parameter of the latching clearance is successfully completed.

[0222] Optionally, the method further includes:

[0223] S3, storing the design scheme of the new design requirement as a design example in the tolerance knowledge instance library. This kind of design enriches the instance library and improves the production efficiency.

[0224] Please refer to Figure 5 , the second aspect of the present invention discloses an instance-based tolerance optimization system for the opening and closing mechanism, and the system includes:

[0225] The first processing module 100 is configured to traverse the tolerance knowledge instance library according to the new design requirement, calculate the requirement similarity between the new design requirement and each design instance, and sort according to the requirement similarity, and screen out K design instances higher than the similarity threshold as similar design instances; wherein, multiple design instances are stored in the tolerance knowledge instance library, and each design instance includes key measurement dimensions, nominal dimensions of parts, and tolerance information; the key measurement dimensions and nominal dimensions of parts belong to the design requirement;

[0226] The second processing module 200 is configured to calculate the assembly accuracy according to the tolerance information in each similar design instance, and then judge whether the tolerance information in any similar design instance meets the new design requirement according to the calculated assembly accuracy; wherein,

[0227] If the tolerance information in the K similar design instances does not meet the new design requirement, perform the following actions:

[0228] Based on the target optimization model of the processing cost and quality loss function, with the process capability index and tolerance accumulation as the constraint conditions, optimize the tolerance information in any similar design instance to obtain a design scheme that meets the new design requirement.

[0229] Optionally, the target optimization model is:

[0230]

[0231] Among them, min represents minimization; C is the objective function; ω i is the weighted coefficient converted according to the contribution degree analysis of part features; T iis the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the part feature tolerance, where c0, c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the part feature quality loss function, where A is the limit loss when the product is within the tolerance, and T is the design tolerance;

[0232] The constraint conditions include:

[0233]

[0234] C pk min ≤C pk ≤C pk max

[0235]

[0236] T i ≤T

[0237] where C Pmin is the lower limit of the process capability index; C Pmax is the upper limit of the process capability index; C Pi is the process capability index corresponding to the i-th tolerance; σ is the standard deviation of the process in a stable state; C Pk min is the lower limit of the actual process capability index; C Pk max is the upper limit of the actual process capability index; C Pk is the actual process capability index, where USL and LSL are the upper and lower limits of the tolerance control requirements, which are the key measurement dimensions in the new design requirements, μ is the mean of the size distribution statistically obtained in the actual machining process; λ i is the weighting coefficient of the i-th tolerance, T 0i is the assembly accuracy of the new design requirements.

[0238] The third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. When the processor executes a computer program stored in the memory, the steps in any one of the methods for optimizing the tolerance of the opening and closing mechanism based on instances in the first aspect of the present disclosure are implemented.

[0239] Figure 6 is a structural diagram of an electronic device according to an embodiment of the present invention, as Figure 6As shown in the figure, the electronic device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, a carrier network, near field communication (NFC), or other technologies. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen. The input device of the electronic device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the electronic device, or an external keyboard, touchpad, or mouse, etc.

[0240] Those skilled in the art can understand that Figure 6 the structure shown in the figure is only a structural diagram of a part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of this application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0241] The fourth aspect of the present invention discloses a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps in a method for optimizing the tolerance of an opening and closing mechanism based on instances in any one of the first aspects of the present disclosure are implemented.

[0242] In summary, the technical solution proposed by the present invention has the following technical effects: On the basis of ensuring that the assembly accuracy is qualified, the method comprehensively considers the processing cost and the process capability index, and designs and optimizes the tolerance of the opening and closing mechanism. Based on the tolerance instance library of the opening and closing mechanism, through the analysis and calculation of the contribution degree and sensitivity, the optimization objectives of the product processing cost and quality loss are constructed, and an objective optimization model with the process capability index and tolerance accumulation as constraint conditions is established, so that the processing cost and the assembly quality loss reach a balance, aiming to meet the design accuracy under precision manufacturing, improve production efficiency, improve part quality, and further reduce production costs.

[0243] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that the technical solutions recorded in the foregoing embodiments can still be modified, or some or all of the technical features can be equivalently replaced, and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An instance-based tolerance optimization method for opening and closing mechanisms, characterized in that The method includes: Step S1: Traverse the tolerance knowledge instance library according to the new design requirements, calculate the requirement similarity between the new design requirements and each design instance, sort according to the requirement similarity, and screen out K design instances with a similarity higher than the similarity threshold as similar design instances; Among them, the tolerance knowledge instance library stores multiple design instances, and each design instance includes key measurement dimensions, nominal dimensions of components, and tolerance information; the key measurement dimensions and nominal dimensions of components belong to the design requirements; Step S2: Calculate the assembly accuracy according to the tolerance information in each similar design instance, and then judge whether the tolerance information in any similar design instance meets the new design requirements according to the calculated assembly accuracy; where: If the tolerance information in the K similar design instances does not meet the new design requirements, the following steps are executed: Based on the objective optimization model of machining cost and quality loss function, with the process capability index and tolerance accumulation as constraint conditions, optimize the tolerance information in any similar design instance to obtain a design scheme that meets the new design requirements.

2. The method according to claim 1, characterized in that In the step S1, the KNN algorithm is used to calculate the requirement similarity between the new design requirements and each design instance.

3. The method according to claim 2, wherein In the step S2, the calculation formula for the assembly accuracy is: Among them, T mi is the calculated assembly accuracy, T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; λ i is the weighting coefficient of the i-th tolerance.

4. The method according to claim 2, wherein In the step S2, the objective optimization model is: Among them, min represents minimization; C is the objective function; ω i is the weighting coefficient converted according to the analysis of the contribution degree of component features; T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the component feature tolerance, where c0, c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the quality loss function of the component feature, where A is the limit loss of the product within the tolerance regulation, and T is the design tolerance.

5. The method according to claim 4, wherein In the step S2, the constraint conditions include: C pkmin ≤C pk ≤C pkmax T i ≤T Among them, C Pmin is the lower limit of the process capability index; C Pmax is the upper limit of the process capability index; C Pi is the process capability index corresponding to the i-th tolerance; σ is the standard deviation of the process in a stable state; C Pkmin is the lower limit of the actual process capability index; C Pkmax is the upper limit of the actual process capability index; C Pk is the actual process capability index, Among them, USL and LSL are the upper and lower limits of the tolerance control requirements respectively, which are the key measurement dimensions in the new design requirements, μ is the mean value of the size distribution statistically in the actual processing process; λ i is the weighting coefficient of the i-th tolerance, T 0i is the assembly accuracy of the new design requirements.

6. The method according to claim 1, characterized in that, The method further includes: S3, storing the design scheme of the new design requirements as a design instance in the tolerance knowledge instance library.

7. An instance-based tolerance optimization system for opening and closing mechanisms, characterized in that, The system includes: The first processing module is configured to traverse the tolerance knowledge instance library according to the new design requirements, calculate the requirement similarity between the new design requirements and each design instance, sort according to the requirement similarity, and screen out K design instances with a similarity higher than the similarity threshold as similar design instances; among them, the tolerance knowledge instance library stores multiple design instances, and each design instance includes key measurement dimensions, nominal dimensions of components, and tolerance information; the key measurement dimensions and nominal dimensions of components belong to the design requirements; The second processing module is configured to calculate the assembly accuracy according to the tolerance information in each similar design instance, and then judge whether the tolerance information in any similar design instance meets the new design requirements according to the calculated assembly accuracy; where, If the tolerance information in the K similar design instances does not meet the new design requirements, the following actions are performed: Based on the objective optimization model of machining cost and quality loss function, with the process capability index and tolerance accumulation as constraint conditions, optimize the tolerance information in any similar design instance to obtain a design scheme that meets the new design requirements.

8. The system according to claim 7, characterized in that The objective optimization model is: where min represents minimization; C is the objective function; ω i is the weighting coefficient converted according to the analysis of the contribution degree of component features; T i is the i-th tolerance value in the tolerance information, and n is the number of tolerances in the tolerance information; C m (T i ) is the machining cost function of the component feature tolerance, where c0 ， c1, c2, c3 are known parameters related to the tolerance; L(T i ) is the quality loss function of the component feature, where A is the limit loss of the product within the tolerance regulation, and T is the design tolerance; The constraint conditions include: C pkmin ≤C pk ≤C pkmax T i ≤ T Among them, C Pmin is the lower limit of the process capability index; C Pmax is the upper limit of the process capability index; C Pi is the process capability index corresponding to the i-th tolerance; σ is the standard deviation of the process in a stable state; C Pkmin is the lower limit of the actual process capability index; C Pkmax is the upper limit of the actual process capability index; C Pk is the actual process capability index, where USL and LSL are the upper and lower limits of the tolerance control requirements respectively, and are key measurement dimensions in the new design requirements, μ is the mean of the size distribution statistically obtained in the actual machining process; λ i is the weighting coefficient of the i-th tolerance, T 0i is the assembly accuracy of the new design requirements.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. When the processor executes the computer program, the steps in any one of claims 1 to 6 of a tolerance optimization method for an opening and closing mechanism based on instances are implemented.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps in any one of claims 1 to 6 for an instance-based tolerance optimization method of an opening and closing mechanism are implemented.

Images