Fuzzy demand driven product customization design method and system

Abstract

The invention relates to the technical field of product design, and discloses a fuzzy demand-driven product customization design method and system, and the method comprises the steps: a user inputs a fuzzy demand and a preference weight value through a human-computer interaction interface, and the system converts the fuzzy demand into a mathematical model which can be recognized by a computer through a built-in logic and algorithm; accessing a database by the system, and performing dimension reduction and instance retrieval based on a neural network model to obtain a product instance which can best meet user requirements; and identifying unsatisfied demand items in the retrieved instance products, and extracting related modules from the database for replacement or variant design to obtain a new product model satisfying user demands. According to the product customization rapid design method and system driven by the fuzzy demand, the new product model meeting the personalized demand is generated through the fuzzy demand and the preference weight input by the user by utilizing algorithm conversion, neural network model retrieval and model replacement or variant design, and the flexibility of product customization and the user satisfaction degree are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of product design, and in particular to a fuzzy demand-driven product customization design method and system. Background Art

[0002] In recent years, the trend of market globalization has promoted the change of mass consumption concepts. Consumers are no longer satisfied with standardized mass products, but instead pursue customized products that can meet personal preferences. In response to this demand, contemporary companies have begun to develop product customization system software platforms, aiming to allow customers to participate in the product design process, in order to control costs while producing customized products that meet customer expectations.

[0003] However, there are two major technical problems with existing enterprise product customization system software. On the one hand, some software allows customers to customize existing products by providing selection tables. Although this method enriches the production experience of enterprises, it is difficult to quickly adapt to changing market demands.

[0004] On the other hand, some software allows customers to input clear parameter requirements to customize new personalized products. However, this not only wastes existing design resources, but also places too high demands on customers. Because most customers are not professional designers and lack understanding of professional design parameters, it is difficult to provide complete and accurate demand information. For customers who can only provide vague semantic requirements, such as "probably", "better", "possible", etc., it undoubtedly further increases the difficulty of product customization. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a fuzzy demand-driven product customization design method and system to solve the above technical problems.

[0006] In the first aspect, a fuzzy demand driven product customization rapid design method is provided, comprising the following steps: Step S1: The user inputs fuzzy requirements and preference weight values ​​through the human-computer interaction interface, and the system converts the fuzzy requirements into a mathematical model that can be recognized by the computer through built-in logic and algorithms; Step S2: the system accesses the database, performs dimension reduction and instance retrieval based on the neural network model, and obtains the product instance that best meets the user's needs; Step S3: Identify the unmet demand items in the retrieved example products, extract relevant modules from the database for replacement or variant design, and obtain a new product model that meets user needs.

[0007] Furthermore, the step S1 includes: Based on the rationality and consistency judgment of association rules, the user input requirements are logically checked and completed; Based on the fuzzy semantic quantification processing of triangular fuzzy numbers, the fuzzy requirements input by users are converted into definite values.

[0008] Furthermore, the step S2 includes: Based on the dimension reduction of the convolutional neural network model, feature extraction and classification of historical order data are performed; Based on the product classification calculated by the distance value, determine the class library containing product instances that are most likely to meet user needs; Based on instance retrieval calculated by cosine value, the product instance most similar to the user's needs is retrieved from the determined class library.

[0009] Furthermore, the distance value is calculated using generalized Euclidean distance.

[0010] Furthermore, the cosine value calculation is used to measure the similarity between user requirements and product instances.

[0011] Furthermore, step S3 includes: Identify the modules corresponding to the unsatisfied requirements in the retrieved instance product; Extract modules that have an impact on unsatisfied requirements from the database and replace them; When there is no replaceable module, similar modules are redesigned by parametric CAD modeling based on the API functions provided by SolidWorks.

[0012] In a second aspect, a fuzzy demand-driven product customization design system is provided, based on a fuzzy demand-driven product customization design method described in any one of the above, comprising: A human-computer interaction interface for receiving fuzzy requirements and preference weight values ​​input by users; Database, which stores historical product orders and corresponding product instances; Processing module, used to convert fuzzy requirements into mathematical models, and perform instance retrieval and variant design; CAD interface module, used to call SolidWorks API functions for parametric modeling.

[0013] Furthermore, the processing module includes: Demand processing submodule, used for demand verification and quantification based on association rules and triangular fuzzy numbers; The instance retrieval submodule performs dimension reduction and instance retrieval based on a neural network model; The variant design submodule is used to identify items that do not meet requirements and perform module replacement or parametric CAD modeling.

[0014] Furthermore, the system also includes a connection module for establishing a connection between the QT software and SolidWorks and MySQL databases to achieve data interaction and model storage.

[0015] The invention adopting the above technical solution has the following advantages: The present invention generates new product models that meet personalized needs through fuzzy needs and preference weights input by users, using algorithm conversion, neural network model retrieval, and model replacement or variant design, thereby improving the flexibility of product customization and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the specific implementation of the present invention, the following will briefly introduce the drawings required for use in the specific implementation. In all the drawings, each element or part is not necessarily drawn according to the actual scale.

[0017] Figure 1 A flowchart of a fuzzy demand-driven product customization design system of the present invention; Figure 2 A flow chart of a fuzzy demand driven product customization design method of the present invention; Figure 3 This is a product classification display diagram in a fuzzy demand-driven product customization design system of the present invention. DETAILED DESCRIPTION

[0018] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0019] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should be the common meanings understood by technicians in the field to which the present invention belongs. The terms "first", "second", etc. in the specification and claims of the embodiments of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to implement the embodiments of the present disclosure described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. Unless otherwise specified, the term "multiple" means two or more. In the embodiments of the present disclosure, the character " / " indicates that the objects before and after are in an "or" relationship. For example, A / B means: A or B. The term "and / or" is a description of the association relationship of objects, indicating that there can be three relationships. For example, A and / or B means: A or B, or, A and B. The term "corresponding" can refer to an association relationship or a binding relationship, and A and B correspondingly refer to an association relationship or a binding relationship between A and B.

[0020] like Figure 1 to Figure 3 As shown, a fuzzy demand-driven product customization design method of the present invention comprises the following steps: Step S1: The user inputs fuzzy requirements and preference weight values ​​through the human-computer interaction interface, and the system converts the fuzzy requirements into a mathematical model that can be recognized by the computer through built-in logic and algorithms; Step S2: the system accesses the database, performs dimension reduction and instance retrieval based on the neural network model, and obtains the product instance that best meets the user's needs; Step S3: Identify the unmet demand items in the retrieved example products, extract relevant modules from the database for replacement or variant design, and obtain a new product model that meets user needs.

[0021] Specifically, through the fuzzy demands and preference weights input by users, new product models that meet personalized needs are generated by utilizing algorithm transformation, neural network model retrieval, and model replacement or variant design, thereby improving the flexibility of product customization and user satisfaction.

[0022] In some embodiments, step S1 includes: Based on the rationality and consistency judgment of association rules, the user input requirements are logically checked and completed; Based on the fuzzy semantic quantification processing of triangular fuzzy numbers, the fuzzy requirements input by users are converted into definite values.

[0023] Specifically, demand acquisition Reasonableness and consistency judgment based on association rules: For general customers, due to lack of professional knowledge, there may be unreasonable phenomena in the demand information they submit. For demand attributes with contradictions or conflicts, the system platform should guide customers to make correct modifications. Association rules refer to the potential connection between attributes, and four rules are set: mandatory relationship, exclusive relationship, optional relationship, and inheritance relationship.

[0024] Based on product features and the opinions of consulting experts, we first determine whether the customer's requirements are consistent with the logical relationship in terms of product structure, function and performance.

[0025] For example, when ordering drone products, the demand for high shooting resolution and high mapping accuracy will inevitably lead to an increase in the budget. The demand for good low-altitude flight performance and large-range altitude flight can be selected at the same time. The demand for extremely high endurance and small size are mutually exclusive. For incomplete demand information, the system platform guides customers to complete it based on association rules.

[0026] Fuzzy semantic quantification processing based on triangular fuzzy numbers: The customer demand information obtained usually contains fuzzy semantics.

[0027] Triangular fuzzy numbers are a method of converting uncertain fuzzy language into definite values. It can solve the contradiction that the described object cannot be accurately measured and can only be described with fuzzy language. Customer fuzzy demand information can be quantified based on triangular fuzzy numbers.

[0028] Customer fuzzy demands are divided into option-based fuzzy demands and parameter-based fuzzy demands. For option-based fuzzy demands, such as "moderate" and "high", a linguistic triangular fuzzy set is established. . It represents an ordered set of language evaluation values, is a language evaluation value in the set. The triangular fuzzy number of the evaluation value can be expressed as:

[0029] in, is the serial number value of this language evaluation object in this set of ordered language evaluation values, The maximum ordinal value in this ordered set of language evaluation values.

[0030] For example, the evaluation value "moderate" is a language evaluation value in the language set (low, lower, moderate, higher, high), is 2, is 4, and its triangular fuzzy number can be expressed as =(0.25,0.5,0.75). The quantization result takes the middle value of the triangular fuzzy number, that is, the final quantization result with the evaluation value of “moderate” is 0.5.

[0031] For parametric fuzzy requirements, such as "about 22" and "about 4.5", triangular fuzzy numbers are used to convert linguistic variables into fuzzy membership, that is, ,in yes The membership distribution function is as follows.

[0032]

[0033] in, is the lower boundary value, is the upper boundary value, is the middle value. A set of "speeds" is defined, where 25 and above are considered "high" speeds; 15 and below are considered "low" speeds; and 15 to 25 are considered "medium" speeds. represents the parametric fuzzy requirements provided by the customer, then The membership function of is as follows.

[0034]

[0035] when When the value is 22, its membership function is equal to 0.6. That is, when the customer proposes a fuzzy demand for "speed" of "22", the quantified result is 0.6.

[0036] For demand attributes with opposing relationships, such as discrete "yes" and "no", and enumeration "red", "yellow", and "green", comparisons and judgments should be made in the software, and if they are different, they should be directly taken as "1" and "0".

[0037] Through the above processing, the system converts the customer's fuzzy demands into a set of simple mathematical models ranging from 0 to 1.

[0038] In some embodiments, step S2 includes: Based on the dimension reduction of the convolutional neural network model, feature extraction and classification of historical order data are performed; Based on the product classification calculated by the distance value, determine the class library containing product instances that are most likely to meet user needs; Based on instance retrieval calculated by cosine value, the product instance most similar to the user's needs is retrieved from the determined class library.

[0039] In some embodiments, the distance value is calculated using generalized Euclidean distance.

[0040] In some embodiments, the cosine value calculation is used to measure the similarity between the user demand and the product instance.

[0041] Specifically, instance retrieval Data processing based on neural network model: The CNN convolutional layer filters the data and extracts the features of the input data.

[0042] It is believed that the customer demand data of historical orders in the current database are the results of screening, which can represent the fuzzy demands raised by the customers this time, that is, there are always product instances in the database that meet the demands.

[0043] The CNN pooling layer reduces the dimensionality of the data while retaining important features of the data.

[0044] When the number of retrieved instances in the database is huge, a small number of instances of equal quantity can be randomly selected from each instance library after the product category division, and the "distance value" between the demand corresponding to each instance and the demand proposed by the customer this time can be calculated. Among the N instances with the closest "distance", which category of products has the largest number? It is considered that the instance library of this category of products has product instances that meet the customer's needs, thereby completing the product classification.

[0045] Product classification based on distance value calculation: In mathematics, a numerical value can be used to describe the "distance" between the two, and then the difference between the two can be described by the "distance". Generally speaking, the smaller the "distance", the smaller the difference and the greater the similarity. Customer demand can be converted into a set of mathematical models. By combining the expert's guidance on the objective importance of demand attributes and the user's own subjective preference requirements, the importance of demand attributes is weighted. By accessing the database, the first The generalized Euclidean distance between the instance sample and the sample to be designed :

[0046] in, Indicates the sample to be designed Requirement attributes; Indicates Example sample Requirement attributes; Indicates The weight value corresponding to each requirement attribute; Represents the requirement attribute dimension.

[0047] like Figure 3 As shown, taking the two-dimensional plane diagram as an example, assuming that the maximum dimension of product demand is two dimensions, the black and gray spheres are the coordinate distribution of the quantified demand for the first and second categories of products, respectively, and the white spheres are the coordinate positions of the quantified demand proposed by the customer this time. Taking the 8 spheres closest to the white sphere, it can be seen that the gray spheres are the majority, that is, the white spheres are more similar to the second category of products. The product instances that best meet customer needs can be retrieved from the second category product instance library, thereby achieving the purpose of reducing the dimension.

[0048] The CNN fully connected layer integrates data and regresses features.

[0049] Calculate the "cosine value" of the similarity between the requirements corresponding to each instance in the target instance library and the requirements raised by the customer this time, and take the product instance with the cosine value closest to 1 as the final retrieval instance.

[0050] Instance retrieval based on cosine calculation: Cosine similarity measures the similarity of objects by calculating the cosine value of the angle between the vectors from the center of the circle to the two points, and focuses more on the difference in direction between the two points rather than their positions.

[0051] Using cosine similarity, the range of cosine value is between -1 and 1. The larger the cosine value, the smaller the vector angle, and the higher the similarity between the two. When the cosine value is 1, the two vectors coincide, and it can be said that the things represented by the two vectors are completely similar. The cosine similarity test can be derived using the following formula:

[0052] in, For the sample to be designed Requirement attributes; For the Example sample Requirement attributes; It is the demand attribute dimension.

[0053] In some embodiments, step S3 includes: Identify the modules corresponding to the unsatisfied requirements in the retrieved instance product; Extract modules that have an impact on unsatisfied requirements from the database and replace them; When there is no replaceable module, similar modules are redesigned by parametric CAD modeling based on the API functions provided by SolidWorks.

[0054] Specifically, the variant design Secondary development based on CAD technology: The retrieved example products should be evaluated on product parameter performance by an expert team composed of professional technicians based on the product customization system software platform.

[0055] When customer needs are not fully met, the system identifies the modules that do not meet the requirements after product module division, and takes out other modules from the database for replacement and reorganization. When there are no suitable modules in the database, similar models are taken out for parametric CAD modeling and redesign. That is, based on the relevant design parameters, the system will automatically generate new products based on the existing parametric models and dimensional constraints, thereby realizing rapid product variation design. Users can also choose to input design parameters themselves to achieve the purpose of participating in the design.

[0056] SolidWorks uses the Windows operating system as its platform and provides hundreds of application programming interfaces (APIs) based on Object Linking and Embedding (OLE). These interfaces encapsulate all the functions of SolidWorks. The system can implement all operations on SolidWorks, including module modification and reorganization, by calling the API based on Visaul C++.

[0057] The basic idea of ​​realizing product variant design through SolidWorks secondary development based on QT software: Establish the connection between QT software and SolidWorks and MySQL; Take out the 3D model of the design part from the database and load it into SolidWorks; Read the design parts storage data in the database: the mapping relationship that affects the required parameters, determine the parameters that need to be changed to generate different specifications of design parts, and replace them with variables; The newly generated three-dimensional model and related information are stored in the database.

[0058] In some other embodiments, a fuzzy demand-driven product customization design system is provided, and a fuzzy demand-driven product customization design method based on any one of the above items includes: A human-computer interaction interface for receiving fuzzy requirements and preference weight values ​​input by users; Database, which stores historical product orders and corresponding product instances; Processing module, used to convert fuzzy requirements into mathematical models, and perform instance retrieval and variant design; CAD interface module, used to call SolidWorks API functions for parametric modeling.

[0059] In some embodiments, the processing module includes: Demand processing submodule, used for demand verification and quantification based on association rules and triangular fuzzy numbers; The instance retrieval submodule performs dimension reduction and instance retrieval based on a neural network model; The variant design submodule is used to identify items that do not meet requirements and perform module replacement or parametric CAD modeling.

[0060] In some embodiments, the system also includes a connection module for establishing a connection between the QT software and SolidWorks and MySQL databases to achieve data interaction and model storage.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A fuzzy demand driven product customization design method, characterized in that: The following steps are involved: Step S1: The user inputs fuzzy requirements and preference weight values ​​through the human-computer interaction interface, and the system converts the fuzzy requirements into a mathematical model that can be recognized by the computer through built-in logic and algorithms; Step S2: the system accesses the database, performs dimension reduction and instance retrieval based on the neural network model, and obtains the product instance that best meets the user's needs; Step S3: Identify the unmet demand items in the retrieved example products, extract relevant modules from the database for replacement or variant design, and obtain a new product model that meets user needs.

2. A fuzzy demand driven product customization design method according to claim 1, characterized in that: The step S1 includes: Based on the rationality and consistency judgment of association rules, the user input requirements are logically checked and completed; Based on the fuzzy semantic quantification processing of triangular fuzzy numbers, the fuzzy requirements input by users are converted into definite values.

3. A fuzzy demand driven product customization design method according to claim 2, characterized in that: The step S2 includes: Based on the dimension reduction of the convolutional neural network model, feature extraction and classification of historical order data are performed; Based on the product classification calculated by the distance value, determine the class library containing product instances that are most likely to meet user needs; Based on instance retrieval calculated by cosine value, the product instance most similar to the user's needs is retrieved from the determined class library.

4. The fuzzy demand-driven product customization design method according to claim 3, characterized in that: The distance value is calculated using generalized Euclidean distance.

5. The fuzzy demand-driven product customization design method according to claim 3, characterized in that: The cosine value calculation is used to measure the similarity between user requirements and product instances.

6. The fuzzy demand-driven product customization design method according to claim 1, characterized in that: The step S3 includes: Identify the modules corresponding to the unsatisfied requirements in the retrieved instance product; Extract modules that have an impact on unsatisfied requirements from the database and replace them; When there is no replaceable module, similar modules are redesigned by parametric CAD modeling based on the API functions provided by SolidWorks.

7. A fuzzy demand driven product customization design system, based on a fuzzy demand driven product customization design method according to any one of claims 1 to 6, characterized in that: include: A human-computer interaction interface for receiving fuzzy requirements and preference weight values ​​input by users; Database, which stores historical product orders and corresponding product instances; Processing module, used to convert fuzzy requirements into mathematical models, and perform instance retrieval and variant design; CAD interface module, used to call SolidWorks API functions for parametric modeling.

8. The system according to claim 7, characterized in that The processing module comprises: Demand processing submodule, used for demand verification and quantification based on association rules and triangular fuzzy numbers; The instance retrieval submodule performs dimension reduction and instance retrieval based on a neural network model; The variant design submodule is used to identify items that do not meet requirements and perform module replacement or parametric CAD modeling.

9. The system according to claim 8, characterized in that The system also includes a connection module for establishing a connection between the QT software and SolidWorks and MySQL databases to achieve data interaction and model storage.

Images