CN116484525A - Feature mapping-based reverse generation method and system for process model - Google Patents

Abstract

The invention discloses a process model reverse generation method and system based on feature mapping, comprising the following steps: S1: building a feature mapping library; S2: extracting topological features of the surface of parts; S3: converting design features based on the feature mapping library to processing features; S4: generating a three-dimensional process model. The geometric feature-based method of the present invention can be applied to various types of parts, and more types of geometric features can be identified; the geometric features of the parts can be calculated more accurately, avoiding the error accumulation problem that may occur in the rule-based method; The efficiency of recognition is improved; new geometric feature recognition methods can be easily added, thereby expanding the function and scope of application of the algorithm.

Description

Method and system for reverse generation of process model based on feature mapping

technical field

The invention relates to the technical field of computer-aided process design, in particular to a method and system for reverse generation of a process model based on feature mapping.

Background technique

The manufacturing industry has always been the core driving force for the country's progress, and it is also an area where my country needs to make key breakthroughs in promoting modern production. With the continuous advancement of computer and automatic control technology, the overall processing efficiency and processing quality of my country's manufacturing industry have been greatly improved. Of course, innovation and development must be persisted in multiple links in the entire cycle of product design and manufacturing.

With the rapid development of science and technology, the manufacturing industry has begun to develop in the direction of flexibility and integration. Computer Aided Process Planning (CAPP) is the core technology to realize the integration of product design and manufacturing, and 3D CAPP is the direction of future manufacturing development. The part model file defined by 3D technology has the advantages of complete model information, model visualization and model parameter information. The use of 3D CAPP can ensure the scientific nature of process planning and complete the planning with high quality and efficiency, and realize the integration of product design and manufacturing. The 3D process model generation technology is the key technology to realize 3D CAPP. The current method of generating 3D process model is mostly completed by process designers after they understand the completed part design entity model, and then complete the human-computer interaction modeling. The construction of the 3D process model in this way greatly increases the workload of the process planning staff and also affects the quality of the process planning, and because the designer and the craftsman have different understandings of the part model, there is a certain degree of disconnection between design and processing.

The construction of 3D process model oriented to manufacturing is the key to complete 3D CAPP. The 3D process model is used as the information basis of the manufacturing process to achieve more coherent data in the product design and manufacturing process and improve processing efficiency and quality. The three-dimensional process model is to express various process information such as the processing method and processing parameters related to the current process of the part in the form of a three-dimensional entity. The generation of the three-dimensional process model is to realize the part surface feature recognition with the help of the computer, which is an important part of the CAPP system.

The current process model generation method lacks discussion on the correlation between each process model, and the function of the 3D process model to transmit manufacturing information has not been fully developed. It cannot be used as the information basis for the entire production cycle of the product, which affects the efficiency of 3D process design. Now that 3D process design is developing towards integration and intelligence, it is necessary to develop a rich feature expert database to realize the rapid development of 3D process.

In view of the above-mentioned defects, the creator of the present invention has finally obtained the present invention through long-term research and practice.

Contents of the invention

In order to solve the above-mentioned technical defects, the object of the present invention is to provide a method and system for reverse generation of process model based on feature mapping.

A process model reverse generation method based on feature mapping, comprising the following steps:

S1: Build a feature map library

According to the requirements, multiple basic features are combined to form a custom feature body, and the custom feature body is stored in the mapping feature library in the form of an attribute adjacency graph;

S2: Extract topological features of part surface

Traverse the design model feature tree and its attribute adjacency relationship to obtain part surface information and generate an attribute adjacency graph;

S3: Conversion of design features to processing features based on feature mapping library

Compare the feature information attribute adjacency graph extracted from the surface of the part with the attribute adjacency graph of the feature mapping library, and find out the internal relationship between the design feature and the processing feature through the feature map, so as to obtain the processing feature and correlate the process parameter information;

S4: Generate 3D process model

Under the guidance of process parameter information, the part design model is reversely restored to the part blank model through feature suppression and each 3D process model is generated. The current 3D process model is created by suppressing the relevant features of the next 3D process, so as to obtain a complete 3D process model.

Further, step S1 specifically includes the following content:

Input the processing and manufacturing features, use the custom feature module to integrate multiple basic features into a feature body according to the needs of users, obtain various forms of custom features, and store the custom features in the feature map library in the form of attribute adjacency graphs.

Further, the part surface information in step S2 includes topological information of faces, rings, and edges, and the topological information of faces, rings, and edges is classified according to classification rules.

Further, the classification rules of the topological information of the faces, rings and edges are as follows:

Surface classification: start searching from the surface with the largest surface area, and use the breadth-first algorithm to search for the continuity and adjacency characteristics of the shared edges of adjacent surfaces. First, the surface is divided into thin plate surfaces and trimmed surfaces, and then the thin plate surface is further divided into web surfaces, bending surfaces, inner ring surfaces, wall surfaces, stagnant surfaces, and inner and outer surfaces according to the adjacency characteristics around the surfaces;

Ring classification: according to the type of surface containing the ring, the type of edge it contains, and combined with the breadth-first search algorithm, the ring is divided into outer ring, rib ring, shallow forming ring, inner ring and hole ring;

Edge Classification: According to the type of the two faces sharing the edge and its own geometry type, the edge is divided into trimmed straight edge, trimmed non-straight edge, non-trimmed straight edge and non-trimmed non-straight edge.

Further, in step S3, it is determined that the internal correlation between the design feature and the processing feature includes the following two situations:

For a design feature that exists alone, that is, a design feature that does not interact with other features, the processing feature is consistent with the design feature;

For two or more design features, if there is a certain connection between them, that is, the manufacturing process is involved with each other, these design features are integrated into a processing feature.

Further, in step S3, the design features are transformed into processing features, specifically including the following:

Firstly, the type of a certain design feature is reflected in the design model feature tree, and the sequence of creating the design model feature in the modeling and the topological shape information of the design model feature surface are pre-defined;

Then organize and define according to the topological shape information of the feature surface, then iteratively visit the design features, judge the type of design features and the correlation information between design features, and further generate the attribute adjacency graph of the part area features;

Finally, the feature mapping library obtains the processing features by matching features, and then extracts the processing features from the output results.

Further, the feature map in step S3 specifically includes the following:

Recursively access the design model feature tree, obtain design features and judge the type of design features;

Process the geometric information of the feature surface of the design model, and generate the attribute adjacency graph of the feature of the part area;

Compare the attribute adjacency graph with the sub-features in the mapped feature library;

If the comparison is successful, the design feature recognition in the corresponding attribute adjacency graph is converted into a processing feature;

Judging whether the feature traversal is completed, output the result when it is completed, and obtain the process model of the process, if it is not completed, continue to visit the feature tree of the design model.

Further, the expression of the modeling process of the three-dimensional process model in step S4 is:

Among them, DesignModel represents the part design model, WorkBlank represents the part blank, F _ij represents the jth design feature removed in the i-th processing process of the part, n is the total number of processes, and m _i is the total number of features that need to be removed in the i-th processing process.

Further, the relationship between two adjacent three-dimensional process models during feature suppression in step S4 is:

Among them, M _i represents the three-dimensional process model in the middle of the part. The three-dimensional process model M _i plus the features removed by the i-1th process is M _i-1 . The model M i- ₁ corresponding to the i-1th process can be obtained by suppressing the features of M _i .

A system adopting the above-mentioned reverse generation method of a process model based on a feature map, comprising:

Feature mapping library building block: used to combine multiple basic features to form a custom feature body according to requirements, and store the custom feature body in the mapping feature library in the form of an attribute adjacency graph;

Part surface topological feature extraction module: used to traverse the design model feature tree and its attribute adjacency relationship to obtain part surface information and generate attribute adjacency graph;

Processing feature acquisition module: it is used to compare the feature information attribute adjacency graph extracted from the surface of the part with the attribute adjacency graph of the feature mapping library, and find out the internal relationship between design features and processing features through feature mapping, so as to obtain processing features and correlate process parameter information;

3D process model generation module: under the guidance of process parameter information, the part design model is reversely restored to the part blank model through feature suppression and each 3D process model is generated, and the current 3D process model is created by suppressing the relevant features of the next 3D process, so as to obtain a complete 3D process model.

Compared with the prior art, the beneficial effects of the present invention are:

1. Wide applicability: The rule-based method in the prior art needs to set rules first, and these rules are usually only applicable to specific types of parts or specific geometric features, and cannot adapt to complex and diverse situations, while the method based on geometric features of the present invention can be applied to various types of parts, and can identify more types of geometric features;

2. High precision: The method based on geometric features can calculate the geometric features of parts more accurately, avoiding the error accumulation problem that may occur in rule-based methods;

3. High degree of automation: The method based on geometric features can automatically identify the geometric features of parts without manual setting of rules, thus improving the efficiency of recognition;

4. Good scalability: The method based on geometric features can easily add new geometric feature recognition methods, thereby expanding the function and scope of application of the algorithm.

Description of drawings

Fig. 1 is a schematic diagram of the automatic generation flow chart of the process model of the present invention;

Fig. 2 is a schematic diagram of the classification process of the rings of the present invention;

Fig. 3 is a schematic diagram of the internal mapping relationship between the design feature and the processing feature of the present invention;

Fig. 4 is a feature mapping flowchart of the present invention;

Fig. 5 is a flow chart of adding and constructing the processing feature map library of the present invention;

Fig. 6 is the flow chart of process model generation of the present invention;

Fig. 7 is a schematic diagram of the parametric design of the roof body of the present invention.

Detailed ways

The above and other technical features and advantages of the present invention will be described in more detail below in conjunction with the accompanying drawings.

In the prior art, a rule-based method is disclosed for feature recognition, but the types of features for recognition research are not wide enough, and the output mode of feature parameters and the relationship between features is relatively primitive, limited to features related to simple deformation, and cannot adapt to complex features that may appear. The present invention performs feature recognition based on geometric features, that is, abstracts the topological relationship between part features into the form of nodes and edges to form an attribute adjacency graph. Each node represents a part feature, and each edge represents the topological relationship between part features. Compared with the rule-based method, the geometric feature-based method has the advantages of wider applicability, higher precision, higher degree of automation, and better scalability, and the invention proposes that the process model reverse generation method is realized based on feature mapping, that is, the feature information extracted from the part surface is identified, compared with the attribute adjacency graph of the sub-feature connection stored in the feature map library, the identified design features are converted into processing features, and a construction method of the processing feature map library is provided, and then the processing features are obtained through feature mapping, and then the part design model is reversely restored through processing feature suppression Model the blank of the part and generate each 3D process model to obtain a complete 3D process model.

Example 1

In this embodiment, a process model reverse generation method based on feature mapping includes the following steps:

S1: Build a machining feature mapping library

This method builds a mapping feature library on the basis of the user-defined feature module of UGNX software. User-defined features can be customized according to user requirements. This method system is a plug-in for secondary development on the UGNX platform. The design principle of the user-defined feature module is to integrate multiple basic features into a feature body according to the needs of operating users, so that many forms of features can be customized and stored in the mapping feature library in the form of an attribute adjacency graph. Feature bodies can be defined for features common to the type of part being processed.

As shown in Figure 5, the processing feature is input, the feature body is constructed by using the custom module, and the attribute adjacency graph is generated through the tool of the processing feature management library, and stored in the mapping feature library.

S2: Extract topological features of part surface and construct attribute adjacency graph

The automatic generation of the 3D process model is based on the 3D design model as the only information basis, and the surface information of the design model needs to be obtained. This method is based on the UGNX platform, so it is only necessary to traverse the design model feature tree and its attribute adjacency relationship to obtain the part surface information, and then generate the attribute adjacency graph of the part.

Surfaces, rings, and edges are the basic elements of part geometry, and the relationship and features between them directly affect the properties and design process of parts. By classifying and analyzing these elements, it is possible to better understand the geometry of the part, its constraints, and their interrelationships with other attributes. This helps to build more accurate and targeted attribute adjacency graphs. The reclassification of topological information such as faces, rings, and edges is the key to constructing more accurate and targeted attribute adjacency graphs. The reason is that in the process of forming design features through operations such as pruning and deformation of the model, faces, rings, and edges with different properties will be generated, which are closely related to the features.

The following are the classification rules for faces, rings and edges:

1. Noodle classification

Classify the faces starting from the face with the largest surface area. For adjacent faces, according to the continuity and adjacency characteristics of their common edges, the breadth-first algorithm is used to search, and the faces are classified according to the following order and rules to provide information for subsequent feature recognition.

(1) Division of thin plate surface and trimmed surface

The front and back of the untrimmed model are thin sheet surfaces (sheet_face, denoted as fsheet); the trimmed surface (trim_face, denoted as f _trim ) is the surface generated by the trimming operation. Adjacent sheet faces are G1 continuous, but not G1 continuous between sheet faces and trimmed faces.

(2) Thin plate subdivision

In this step, the surface of the thin plate needs to be divided into the following types in detail according to the adjacency characteristics around the surface:

Web face: the thin plate face with the largest surface area;

Bending face: the sheet face adjacent to the outer ring of the web face;

Inner ring surface: the thin plate surface adjacent to the web surface but not the outer ring, and the thin plate surface adjacent to the inner ring surface and not trimmed is also the inner ring surface;

Inner and outer sides: According to the classified web surfaces, surrounding bending surfaces and inner ring surfaces included in the inner ring, find the sum of the areas of the front and back sides. The side with the smaller area is the inner side, and the larger one is the outer side. The division of inner and outer surfaces is conducive to improving the efficiency of surface classification operations. This study conducts surface classification and feature recognition based on the inner surface.

Wall surface and stagnant surface: For the thin plate surface that has not been classified, if it is only adjacent to one bending surface, it is called a wall surface; if more than one, it is called a stagnant surface. A stagnant surface is a transitional surface that is subsequently transformed into a connection surface, bend surface or wall surface.

Similar to the web face, the sheet face adjacent to the outer ring of the wall is called the bent face; while the sheet face adjacent to the non-outer ring of the wall is called the inner ring face, and the non-trimmed face adjacent to the inner ring face is also the inner ring face. All thin plate faces are thus classified as web faces, wall faces, stagnation faces, bend faces and inner ring faces.

2. Classification of rings and edges

(1) Classification of rings

The rings in the part model represent the inner and outer boundaries of the surface. The original types are divided into outer rings, inner rings and other three types, which cannot meet the requirements of the recognition features. Therefore, the rings are reclassified according to the recognized features, which conforms to the rules of feature recognition. The classification process of the rings is shown in Figure 2.

The rings are defined as follows:

Outer ring: a subtype of ring extracted by the model, which determines the outer boundary of the face;

Rib ring: It is the inner ring related to the outer boundary of the rib feature. The surrounding inner ring surface is not adjacent to the trimming surface. If it does not contain a plane, or only contains one plane when the depth is 4;

Shallow forming ring: It is an inner ring related to the outer boundary of the shallow forming feature, and the surrounding inner ring surface is not adjacent to the trimming surface, and it is not a rib ring;

Inner ring: is the inner ring related to openings and punched openings, composed of edges associated with distinguishable curves, and contains more than one directed edge;

Hole ring: It is related to the hole or stamping opening feature, it is composed of edges associated with indistinguishable curves, and the number of directed edges is one, which is a special inner ring.

(2) Classification of edges

In order to represent the relationship between features, identify some features and geometric structures as features, the edges need to be reclassified. Edges are reclassified into the following four types according to the types of the two faces sharing the edge and their own geometry type:

1) Trimming straight line edge: it is shared by the thin plate surface and the trimming surface, and its geometric type is straight line;

2) Trim non-straight edge: it is shared by the thin plate surface and the trimmed surface, and the geometry type is curve;

3) Non-trimmed straight line edge: it is shared by two thin plate faces, and the geometry type is straight line;

4) Non-trimmed non-straight line edge: it is shared by two thin plate surfaces, and the geometry type is curve.

S3: Conversion of design features to processing features based on feature mapping library

Compare the part attribute adjacency graph obtained in the previous step with the attribute adjacency graph of the feature mapping library, find out the internal relationship between design features and processing features through feature mapping, convert design features into processing features, and associate process parameter information.

1. Definition and classification of processing features

The processing feature is formed through a series of forming processes and has a certain geometric topological shape of practical significance. The actual meanings represented by the surface features of parts on different occasions will also be different, and the functions expressed will also be different. In the field of design and manufacturing, it is divided into design features and processing features.

Machining features reflect a certain engineering language from the perspective of manufacturing, and are mainly oriented to the geometric topology information of the surface of the part and the machining process, and the machining features are associated with machining methods, machine tools, tools and other processing parameters.

This embodiment is aimed at the characteristics of aerospace roof parts, and the classification of processing features is shown in Table 1.

Table 1: Classification of machining features

2. Mapping from design features to processing features

The morphological features expressed by the design model are not easy for craftsmen to interpret. The transformation of geometric information between the design model and the process system requires the mapping of design features to processing features. Under feature mapping, the design features are transformed into processing features based on manufacturing and technology, that is, the relationship between the two is established, so as to obtain the processing features corresponding to the design features, and then obtain the process model of the processing features from the feature mapping library. The process model includes the processing characteristics, processing methods and processing parameters of each process, and the process of processing is the process of material removal. Therefore, the process model of the process can be generated by adding the material removed in the process to the design model. Therefore, the mapping from design features to processing features is the basis for further realization of process model generation.

Find out the internal relationship between design features and processing features through feature mapping. There are many intersecting features in design features. Design features are the description of this feature, while processing features are more concerned about how to process this feature. In many cases, it is to simplify the problem. Sometimes it is also necessary to take into account the surrounding feature information.

Analysis of the internal relationship between design features and processing features, including the following two situations:

(1) For the design features that exist alone and do not interact with other features, they are basically consistent with the processing features.

(2) For two or more design feature information, if there is some connection between them and the manufacturing process is involved with each other, these design features can be integrated into a processing feature.

Specifically, as shown in Figure 3, the process of machining a part is the process of removing material, such as a boss feature. To process a boss feature, it is necessary to remove the concave material around the boss, that is, to process a concave feature. Thus, the boss is mapped to the concave information, that is, the cavity feature.

The basic process of the feature mapping algorithm for converting design features into processing features is: when the designer uses 3D modeling software to create a model, the type of a certain feature is reflected in the design feature tree, such as the convexity and concaveness of the surface shape. The topological shape information of the feature surface, through the organization and definition of the surface information, can cyclically visit the design features, and judge the type of features and the correlation information between the features, and further generate the attribute adjacency graph of the part area features, use the mapping feature library to construct and obtain the processing features by matching features, and then extract the process model of the processing features from the output results.

As shown in Figure 4, the specific process of feature mapping is as follows:

Recursively access the design model feature tree, obtain design features and judge the type of design features;

Process the geometric information of the feature surface of the design model, and generate the attribute adjacency graph of the feature of the part area;

Compare the attribute adjacency graph with the sub-features in the mapped feature library;

If the comparison is successful, the design feature recognition in the corresponding attribute adjacency graph is converted into a processing feature;

Judging whether the feature traversal is completed, the result is output when it is completed, and the output result is the process model.

S4: Generate 3D process model

From the blank of the part material to the finished part, after the completion of each machining process, a three-dimensional product will be obtained. The change in the surface shape and properties of this three-dimensional product is the process of generating processing features, and vice versa is the process of suppressing processing features, returning to the state of the model before processing. In this paper, under the guidance of the process model obtained in the previous step, the material cut off in this process of the part design model is added to the model through feature suppression. Step by step reverse recovery to the part blank model and generate each 3D process model, so as to obtain a complete 3D process model, the principle is shown in Figure 6.

In the reverse generation process of the 3D process model of parts, the creation of the current 3D process model is realized by suppressing the related features of the next 3D process. Use DesignModel to represent the part design model, WorkBlank to represent the blank of the part, F _ij to represent the jth design feature removed in the i-th processing process of the part, n to be the total number of processes, and m _i to be the total number of features to be removed in the i-th processing process. Then the part modeling process can be expressed as:

Use M _i to represent the 3D process model in the middle of the part, M _n is obtained after suppressing the relevant features on the part design model, and the blank model is obtained after the feature suppression of the 3D process model _M1 . The three-dimensional process model M _i plus the features removed by the i-1th process is M _i-1 , and the model M i- ₁ corresponding to the i-1th process can be obtained by suppressing the features of M _i . The relationship between two adjacent processes is:

Basic principle: the present invention first establishes a feature mapping library based on various knowledge, which is stored in the form of an attribute adjacency graph, then obtains topological information from the feature tree of the part design model and generates an attribute adjacency graph of the part, then performs subgraph matching between the attribute adjacency graph of the part and the attribute adjacency graph of the feature mapping library, and maps the design features to the processing features. , and finally form the rough model, and generate each three-dimensional process model.

Example 2

This embodiment takes the top cover of a typical aerospace part as an example to verify the effectiveness of the system function. There are a large number of top covers in equipment development, with many varieties and small batches. Therefore, using the process model reverse generation method based on feature mapping for 3D process model generation can greatly improve production efficiency. The following is an example process of running the program:

Step1: Start UGNX6.0 and create a new model file;

Step2: Click the 3D top cover parametric design button, and the top cover parametric design window will pop up, as shown in Figure 7;

Step3: Input the key dimension parameters of the roof body, click the OK button to get the 3D model of the roof body;

Step4: Click the process information, import it to generate process model button, and the process route and process information window will pop up. The process information includes the current process name, process number, process description, process personnel, creation time, etc. The process information can be imported in the background or manually input in an interactive way.

The process model generation process is guided by the process information to realize the suppression of the corresponding processing characteristics, and then obtain the corresponding process model. Clicking the process model generation will pop up the feature suppression window, which contains the features that need to be processed in this process, and the processing features correspond to the design parameters and related attributes. The 3D process model of each process link in the process route is generated by processing the 3D roof body parts with feature suppression.

Example 3

In this embodiment, a system of a process model reverse generation method based on feature mapping is provided, including:

Feature mapping library building block: used to combine multiple basic features to form a custom feature body according to requirements, and store the custom feature body in the mapping feature library in the form of an attribute adjacency graph;

Part surface topological feature extraction module: used to traverse the design model feature tree and its attribute adjacency relationship to obtain part surface information and generate attribute adjacency graph;

Processing feature acquisition module: it is used to compare the feature information attribute adjacency graph extracted from the surface of the part with the attribute adjacency graph of the feature mapping library, and find out the internal relationship between design features and processing features through feature mapping, so as to obtain processing features and correlate process parameter information;

3D process model generation module: under the guidance of process parameter information, the part design model is reversely restored to the part blank model through feature suppression and each 3D process model is generated, and the current 3D process model is created by suppressing the relevant features of the next 3D process, so as to obtain a complete 3D process model.

To sum up, the present invention establishes a relationship between the design features in the design model and the specific geometric shape generated by the machining process method through the definition and mapping of the machining features, and generates associated information. Through feature suppression, the product 3D model is reversely restored to the product blank model and each 3D process model is generated to obtain a complete 3D process model.

The present invention has the following advantages:

Wider applicability: Rule-based methods need to set rules first, and these rules are usually only applicable to specific types of parts or specific geometric features, and cannot adapt to complex and diverse situations. However, the method based on geometric features can be applied to various types of parts, and more kinds of geometric features can be recognized.

Higher precision: The geometric feature-based method can calculate the geometric features of the part more accurately, avoiding the error accumulation problem that may occur in the rule-based method.

Higher degree of automation: The method based on geometric features can automatically identify the geometric features of parts without manual setting of rules, thus improving the efficiency of recognition.

Better scalability: The geometric feature-based method can easily add new geometric feature recognition methods, thereby expanding the function and scope of the algorithm.

The above descriptions are only preferred embodiments of the present invention, and are only illustrative rather than restrictive to the present invention. Those skilled in the art understand that many changes, modifications, and even equivalents can be made within the spirit and scope defined by the claims of the present invention, but all will fall within the protection scope of the present invention.

Claims (10)

1. The reverse process model generation method based on the feature mapping is characterized by comprising the following steps of:

s1: construction of feature mapping library

Combining a plurality of basic features according to requirements to form a custom feature body, and storing the custom feature body into a mapping feature library in the form of an attribute adjacency graph;

s2: extracting part surface topology features

Traversing the feature tree of the design model and the attribute adjacency relation thereof to acquire part surface information and generate an attribute adjacency graph;

s3: conversion from design features to process features based on feature mapping library

Comparing the feature information attribute adjacency graph extracted from the surface of the part with the attribute adjacency graph of the feature mapping library, finding out the internal association of the design feature and the processing feature through feature mapping, thereby obtaining the processing feature and associating the technological parameter information;

s4: generating a three-dimensional process model

Under the guidance of technological parameter information, reverse restoring the part design model into a part blank model through feature suppression, generating each three-dimensional process model, and realizing the creation of the current three-dimensional process model through suppressing the related features of the latter three-dimensional process, thereby obtaining a complete three-dimensional process model.

2. The feature mapping-based process model reverse generation method according to claim 1, wherein the step S1 specifically includes the following steps:

inputting processing and manufacturing characteristics, integrating a plurality of basic characteristics according to the requirements of users to obtain a plurality of types of custom characteristic bodies, and storing the custom characteristics into a characteristic mapping library in a mode of attribute adjacency graphs.

3. The feature map-based process model reverse generation method according to claim 1, wherein the part surface information in step S2 includes topology information of faces, rings, and edges, and the topology information of the faces, rings, and edges is classified according to classification rules.

4. The feature mapping-based process model reverse generation method according to claim 3, wherein the classification rules of the topology information of the faces, rings and edges are as follows:

surface classification: searching from the surface with the largest surface area, searching by adopting a breadth-first algorithm aiming at the continuity and adjacent characteristics of the common edges of the adjacent surfaces, firstly dividing the surface into Bao Banmian and a trimming surface, and then further dividing the sheet surface into a web surface, a bending surface, an inner ring surface, a wall surface, a retention surface, an inner side surface and an outer side surface according to the adjacent characteristics around the surface;

ring classification: dividing the ring into an outer ring, a ribbed ring, a shallow ring, an inner ring and Kong Zhuanghuan according to the surface type and the edge type related to the ring and in combination with a breadth-first search algorithm;

edge classification: the sides are divided into a trimmed straight side, a trimmed non-straight side, a non-trimmed straight side and a non-trimmed non-straight side according to the types of the two sides sharing the side and the geometrical type of the side.

5. The feature map-based process model reverse generation method of claim 1, wherein determining the intrinsic correlation of design features and process features in step S3 comprises:

for a design feature that exists alone, i.e., without interaction with other features, the tooling feature remains consistent with the design feature;

for two or more design features, if there is some relationship between them, i.e., the manufacturing process is tied to each other, then the design features are integrated into a single tooling feature.

6. The feature mapping-based process model reverse generation method according to claim 1, wherein the step S3 of converting design features into processing features specifically comprises the following steps:

firstly, reflecting the type of a certain design feature in a design model feature tree, and predefining the sequence of design model feature creation and the topological shape information of the design model feature surface in modeling;

then organizing and defining according to the topological shape information of the feature surface, then circularly accessing the design features, judging the types of the design features and the associated information among the design features, and further generating an attribute adjacency graph of the part region features;

and finally, acquiring the processing characteristics from the characteristic mapping library by a characteristic matching method, and extracting the processing characteristics from the output result.

7. The feature map-based process model reverse generation method according to claim 1, wherein the feature map in step S3 specifically includes the following:

circularly accessing a design model feature tree, obtaining design features and judging the types of the design features;

processing geometric information of the feature surface of the design model to generate an attribute adjacency graph of the feature of the part area;

comparing the attribute adjacency graph with the sub-features in the mapping feature library;

if the comparison is successful, the design feature identification in the corresponding attribute adjacency graph is converted into a processing feature;

judging whether the feature traversal is completed, outputting a result if the feature traversal is completed, extracting processing features according to the output result, and if the feature traversal is not completed, continuing to access the design model feature tree.

8. The feature map-based process model reverse generation method according to claim 1, wherein the modeling process of the three-dimensional process model in step S4 has the expression:

wherein, design model represents a part design model, workBlank represents a part blank, F _ij The j design features of the parts, which represent the i-th machining process step removal, n is the total number of steps, m _i The total number of features to be removed for the ith process step.

9. The method for generating a process model in reverse direction based on feature mapping according to claim 1, wherein the relationship between two adjacent three-dimensional process models at the time of feature suppression in step S4 is:

wherein M is _i Three-dimensional process model representing part middle, three-dimensional process model M _i The characteristic of the i-1 th procedure is M _i-1 For M _i The characteristic is restrained to obtain the i-1 process pairModel M of stress _i-1 。

10. A system employing the feature map-based process model reverse generation method according to any one of claims 1 to 9, comprising:

the feature mapping library construction module: the method comprises the steps of combining a plurality of basic features according to requirements to form a custom feature, and storing the custom feature into a mapping feature library in the form of an attribute adjacency graph;

part surface topological feature extraction module: the method comprises the steps of traversing a design model feature tree and an attribute adjacency relationship thereof to obtain part surface information and generate an attribute adjacency graph;

the processing characteristic acquisition module is used for: the feature information attribute adjacency graph is used for comparing the feature information attribute adjacency graph extracted from the surface of the part with the attribute adjacency graph of the feature mapping library, and the inherent association between the design feature and the processing feature is found out through feature mapping, so that the processing feature is obtained, and the technological parameter information is associated;

the three-dimensional process model generation module: the method is used for reversely recovering the part design model into a part blank model through feature suppression under the guidance of process parameter information, generating each three-dimensional process model, and realizing the creation of the current three-dimensional process model through suppressing the related features of the latter three-dimensional process, thereby obtaining the complete three-dimensional process model.