KR102465451B1 - Process knowledge push method based on machining characteristics - Google Patents

Abstract

The present invention discloses a process knowledge push method based on processing characteristics. First, in order to express the demand and purpose of the part processing process, a method for generating a process intention model based on the processing characteristic vector expression technology is provided; Then, a hierarchical management method of process knowledge is provided, and it is divided into three layers: a basic information layer, a processing information layer, and an auxiliary information layer, and is related to each component in the process intention model. Next, a method for calculating the similarity of the process intention model is provided to improve the efficiency of process knowledge push. Finally, we implement an accurate push of process knowledge based on the process reuse evaluation method. The proposed process intention model generation method can provide technical support for process decision and process information reuse, and the process knowledge push method can lay the foundation for intelligent design of machining process.

Description

Process knowledge push method based on machining characteristics

The present invention relates to the field of intelligent process design of machined parts, and more particularly, to a method of pushing process knowledge based on processing characteristics.

The characteristic-based CAPP system is widely used by manufacturers, and it creates an MBD process model including a large amount of processing characteristics and stores it in the enterprise model database. Such a process model inserts a large amount of processing and manufacturing information such as processing resources and processing demands, but if these process models are not sufficiently discovered and utilized, a large amount of human and material resources are wasted in the enterprise. The characteristic-based process knowledge push method not only solves the problem of reuse of existing process knowledge well, but also implements intelligent process decisions, which can promote the development of intelligent machining design.

Currently, most of the process push methods adopt the form of process knowledge search, mainly process semantic search and shape search method. However, since most of these methods are based on similar searches on the part side and are not related to process machining characteristics, the retrieved process knowledge cannot be directly applied, and moreover, an interaction screening process must be additionally performed.

Among the process push studies based on characteristics and scenarios, the literature “Li Chun-lei, Morong et al. Machining knowledge display and push of geometric evolution drive [J]. Computer Integrated Manufacturing System, 2016, 22(6): 1434-1446” implements push of process knowledge by building a complex network model of machining knowledge, but the above document does not consider the evaluation of machining intention and push knowledge. , it has the disadvantage of not being able to push precisely the required process knowledge. Another document “Zhang Paping (張發平), Li Li (李麗). A study on the method of pushing business process knowledge based on a multi-dimensional hierarchical scenario model [J]. The Journal of Computer-Aided Design and Geometry, 2017, 29(4): 751-758.” proposed a scenario-based knowledge matching and push method, but in the literature, the correlation between process knowledge and scenario models and evaluation of pushed process knowledge There is no explanation for Another document “Sunfu (孫璞), Hou Junje (候俊杰), etc. A study of knowledge push methods for 3D process design [J]. Manufacturing Automation, 2016, 38(9): 96-104” proposed a matching method between process design intent and process knowledge, pushing the corresponding process knowledge through the acquired candidate process knowledge set, but there is no process intention model in the literature. Because it is still semantic description-centric, it cannot be fully utilized for 3D machining design.

In summary, the reuse of process knowledge is already considered by manufacturers as an important factor in shortening the research and production cycle, reducing costs and enhancing corporate competitiveness. However, the process semantic-based search method can be applied only to the low-dimensional process knowledge expression process. The shape-based search method can solve the matching problem of diversified process knowledge, but specific process design requirements are not considered. Therefore, the process knowledge push method based on the machining characteristics can not only be fully inserted into the 3D machining process design system, but also realize high-efficiency reuse of process knowledge.

An object of the present invention is to provide a process knowledge push method based on process characteristics, a process intention model-based similarity calculation and process knowledge evaluation method in order to consider the problems of the prior art and accelerate the development and application of intelligent design for machining processes By doing so, it is to implement the precise push of machining knowledge, improve the efficiency of process design, and provide technical support for intelligent process design.

In order to solve the above technical problem, the present invention provides a process knowledge pushing method based on processing characteristics, which is characterized by including the following steps.

Step (1): Build a process intention model PPIM based on the processing characteristic vector expression method, where the expression of PPIM is PPIM={PPB, PPG, API}, where PPB is the process design background, and PPG is the process design target and API is process accessory information.

Step (2): The similarity calculation method in the process intention model constructed in step (1) includes the calculation of the similarity of basic information of parts, the calculation of the similarity of machining characteristics, and the calculation of the similarity of the quality information.

Step (3): Build a process knowledge stratified representation model, and the process knowledge stratified representation model includes three layers: a basic process information layer, a processing information layer, and a quality information layer.

Step (4): In the process knowledge evaluation method based on the reliability calculation, an optimal process knowledge element is acquired by using the adjacent process attribute reliability calculation value.

Step (5): Implement accurate push of process knowledge based on the part process intention model obtained in step (1) and the reliability value in step (4).

Furthermore, the construction of the process intention model during step (1) is built according to the part design information and processing requirements, where PPB is the process design background including the part type, blank type, material properties and processing type, PPG is a process design target including machining characteristic type, tool feed direction, characteristic surface matrix, proximal surface matrix, topological relationship matrix and specific dimensional information, and API is process accessory information including geometrical precision information and dimensional precision information.

Furthermore, the three layers of the process knowledge layered expression model are specifically as follows.

The basic process information layer includes four process elements: part type, blank type, material properties and processing type.

The machining information layer includes six process elements, namely machining type, machining method, machine tool type, tool type, fixture type and cutting fluid.

The quality information layer contains specific detection requirements.

Furthermore, the specific steps of the process knowledge evaluation method based on the reliability calculation are as follows.

Step (4.1): Prioritize each element of process knowledge.

Step (4.2): Calculate the reliability of the adjacent process knowledge element, and output the most reliable process knowledge element.

Step (4.3): Calculate the reliability of the next step process knowledge elements until all process knowledge elements are obtained.

Furthermore, the PPB is obtained through the generated basic design information, and the PPG is obtained through identification processing characteristics, specifically as follows.

Machining property types are mainly determined by a predefined property library including hole properties, slot properties, planar properties and protrusion properties.

The tool feed direction is determined based on the normal vector of the machining surface.

The characteristic surface matrix is formed through each characteristic surface type and its adjacent surface properties. The characteristic surface types include planar, cylindrical, chamfered, spherical, and circular surfaces. , which includes a concave edge, a protruding edge, and a tangent edge.

The topological relationship matrix is determined by the interrelationships between feature planes, including parallel, perpendicular, oblique and tangent.

The basic dimension information is determined by the minimum bounding box based on processing characteristics, and mainly includes length, width and height.

Furthermore, the processing characteristic similarity calculation in step (2) includes three types: vector matching degree calculation, matrix matching degree calculation, and attribute value matching degree calculation.

Here, the vector matching degree calculation expression is as follows.

conine(p, q) represents the degree of matching between the vector p and the vector q, and i represents the i-th element of the vector, where the matrix matching degree value may be converted into a vector matching degree calculation.

The method of calculating the degree of matching of the matrix is to first convert the N-th matrix into an N-dimensional vector, and then calculate the degree of matching of the vector to implement the calculation of the degree of matching of the matrix.

The expression for calculating attribute value matching is as follows.

S _a (a ₁ , a ₂ ) indicates the degree of matching between attribute values a ₁ and a ₂ , n indicates the number of elements included in the attribute value, and j indicates the j-th element among attribute values.

Furthermore, in the step (4.1), the priority of each element of the process knowledge refers to the priority relationship of the elements included in the machining information layer, and the order of priority is the machining type, machining method, machine tool type, and tool type. , fixture type and coolant type.

Further, the calculation formula of the reliability of the adjacent process knowledge factor in step (4.2) is as follows.

where S _con <p ₁ , p ₂ > represents the reliability of process knowledge elements p ₁ and p ₂ , Preq (p ₁ ) represents the total quantity of process knowledge elements p ₁ , and Preq(p ₁ ∩p ₂ ) denotes the relevant total number of process knowledge elements p ₁ and p ₂ .

Furthermore, the machining characteristic type is expressed as a character string, the tool traversing direction is expressed as a vector, and a characteristic face set matrix and a topological relationship matrix are generated through attribute type assignment values.

Furthermore, in the step (5), specific steps for implementing accurate push of process knowledge based on the obtained part process intention model and reliability value are as follows.

Step (5.1): Through the process intention model, it is determined whether the process background information matching degree in the retrieved process knowledge meets the requirement, if so, proceed to step (5.2), otherwise, continue until the requirement is met Search.

Step (5.2): Determine whether the process target information matching degree in the process knowledge satisfies the requirement, if so, proceed to step (5.3), otherwise, the search is continued until the requirement is met.

Step (5.3): Determine whether the matching degree of process accessory information in the process knowledge meets the requirement, if so, proceed to step (5.4), otherwise, the search is continued until the requirement is met.

Step (5.4): output a list of process information that meets the requirements, then calculate and obtain a confidence value through the adjacent process attribute reliability, and finally output and push the optimal process knowledge element.

Compared with the prior art, the advantages of the present invention are as follows.

The present invention effectively solves the rapid and accurate push of process information during the 3D machining process by proposing a process knowledge push method as a process knowledge expression using processing characteristics as a vector, further improving the efficiency of process design and intelligent process design Provided technical support for the development and application of

1 is a flowchart of the present invention.
2 is a structural diagram of a diesel engine connecting member according to an embodiment.
3 is a diagram illustrating a result of a process design target of a component according to an exemplary embodiment.
4 is a schematic diagram of a process intention model in which a process sequence to be processed is matched according to an embodiment.

Hereinafter, the present invention will be described in more detail with reference to the accompanying drawings and specific examples.

As shown in FIG. 1 , the present invention provides a process knowledge push method based on processing characteristics, which includes the following steps sequentially.

Step (1): Construction of a process planning intent model (PPIM) based on the processing characteristic vector expression method. The expression of PPIM is PPIM={PPB, PPG, API}, where PPB is the process design background, PPG is the process design goal, and API is the process accessory information.

Step (2): Layered representation of process knowledge. The hierarchical representation model of process knowledge is divided into three layers, namely, the basic process information layer, the processing information layer, and the quality information layer.

Step (3): The similarity calculation method of the process intention model mainly includes the calculation of the similarity of the part basic information, the calculation of the similarity of the processing characteristics, and the calculation of the similarity of the quality information.

Step (4): Process knowledge evaluation method based on reliability calculation. An optimal process knowledge factor is obtained by using the adjacent process attribute reliability calculation value.

Step (5): Hierarchical push of process knowledge. Based on the generated process knowledge base and the process design intent model of parts, layered and accurate push of process knowledge is implemented.

The PPIM is generated according to part design information and processing requirements, where PPB includes part type, blank type, material properties and processing type, and PPG is processing characteristic type, tool feed direction, characteristic surface matrix, proximal surface It includes a matrix, a topological relationship matrix, and specific dimension information, and the API includes geometric precision information and dimensional precision information.

The process knowledge layered expression model is managed by generating a process knowledge base, where the process knowledge elements included in the basic process information layer are the same as the process knowledge in the process design background, and the processing information layer includes the processing type and processing method. , machine tool type, tool type, fixture type and cutting fluid contains the contents of six parts, and the quality information layer contains specific detection requirements.

The process intention model similarity calculation includes a similarity calculation of basic process information, a similarity calculation of processing characteristics, and a similarity calculation of quality information.

The process knowledge evaluation method process based on the reliability calculation is as follows. That is, first, the priority of each element of process knowledge is determined. Then, the reliability of the adjacent process knowledge element is calculated, and the process knowledge element with the highest reliability is output. Finally, the reliability of the next step process knowledge element is calculated until all process knowledge elements are obtained.

The process design background knowledge is acquired through basic design information of parts.

The process design target is obtained through identification machining characteristics, wherein the machining property type is mainly determined by a predefined property library including hole properties, slot properties, planar properties and protrusion properties. The tool feed direction is determined based on the normal vector of the machining surface. The characteristic surface matrix is formed through each characteristic surface type and its adjacent surface properties. The characteristic surface types include planar, cylindrical, chamfered, spherical, and circular surfaces. , which includes a concave edge, a protruding edge, and a tangent edge. The topological relationship matrix is determined by the interrelationships between feature planes, including parallel, perpendicular, oblique and tangent. The basic dimension information is determined by the minimum bounding box based on processing characteristics, and mainly includes length, width and height.

Generation of the process knowledge base is organized based on related rules, where the related rules are divided into one-dimensional related rules and multi-dimensional related rules.

The calculation of the similarity of the processing characteristics includes a vector matching degree calculation, a matrix matching degree calculation, and an attribute value matching degree calculation, wherein the vector matching degree calculation expression is as follows.

conine(p, q) denotes the degree of matching between the vector p and the vector q, i denotes the i-th element of the vector, where the matrix matching degree value can be converted into a vector matching degree calculation, and the attribute value matching degree The calculation expression is as follows.

S _a (a ₁ , a ₂ ) indicates the degree of matching between attribute values a ₁ and a ₂ , n indicates the number of elements included in the attribute value, and j indicates the j-th element among attribute values.

The priority of each element of the process knowledge refers to the priority relationship of the elements included in the machining information layer, and the order of priority is a machining type, a machining method, a machine tool type, a tool type, a fixture type, and a cutting fluid type.

The calculation formula for the reliability of the adjacent process knowledge factor is as follows.

where S _con <p ₁ , p ₂ > represents the reliability of process knowledge elements p ₁ and p ₂ , Preq (p ₁ ) represents the total quantity of process knowledge elements p ₁ , and Preq(p ₁ ∩p ₂ ) denotes the relevant total number of process knowledge elements p ₁ and p ₂ .

The machining type is expressed as a character string, the tool traversing direction is expressed as a vector, and a characteristic face set matrix and a topological relationship matrix are created through attribute type assignment values.

The creation of a process intent model is the key to pushing machining process knowledge. The process intent model consists of three parts: process design background, process design objectives, and process ancillary information. Here, the process design background mainly includes product family, part type, blank type, material type, and processing type, and consists of basic attribute information of the machined part. Taking the diesel engine connecting member shown in Fig. 2 as an example, the process design background of the above part is diesel engine connecting member family part, connecting member of V12, casting blank, 45 ^# steel, NC machining.

The process design goal includes the content of six parts: feature type, tool traverse direction, feature face set matrix, proximal face set matrix, topological relationship matrix and dimensional information. To facilitate expression of process design goals and calculation of similarity, each attribute element is managed in the form of a character or binary assignment value. Here, the assigned values of the machining characteristic type and the characteristic surface attribute are shown in Tables 1 and 2. Among the topological relationships, the binary assignment values of parallel, perpendicular, non-perpendicular and tangent are 0001, 0010, 0011, and 0100. Intersecting Edge Properties The binary assignment values of concave, salient, and tangential edges are 0001, 0010, and 0011, and among the types of intersecting edges, the binary assignment values of straight, arc, and spline curves are 0001, 0010 and 0011. In order to do this, the format of “attribute + type”, for example, “tangent edge-arc”, is adopted, and the corresponding value is obtained by multiplying each element value, for example, the value of “tangent edge-circle”. is 0011×0010=0110, and is expressed as 0 if two planes do not intersect. Taking the part shown in Fig. 2 as an example, based on the characteristic identification technique, the surfaces F1 and F3 are obtained as planes, and F2 and F4 are obtained as chamfered surfaces, and based on the binary assignment value of each attribute element, the generated characteristic surface set The matrix, the topological relationship matrix and the adjacent face set matrix are as shown in FIG. 3 .

Table 1 Character assignment values for types of machining characteristics

Table 2 Binary Allocation Values for Machining Characteristics

The process design intent model is a basis for acquiring a candidate process knowledge set, and a similar process design intent model is obtained from an existing process knowledge database by using a matching method, and then the associated process knowledge is acquired to form a candidate process knowledge set. The process of matching similar design intent models is as follows in detail. In other words, in the similarity calculation of the process design background, according to the principle of “similar parts have similar processes,” parts with similar design intent must have the same process design background. no need to perform In the similarity calculation of the process design target, it includes 6 attribute elements, among which the characteristic type and the tool traversing direction must completely match, so the similarity calculation is only performed for the characteristic surface matrix, the proximal surface matrix, the topological relationship matrix and the dimensional information. should be done Based on the company's existing process model database, the process model matching the part slot machining process shown in FIG. 2 and its similarity calculation value are as shown in FIG. 4 . According to the matched process model, related process knowledge is acquired, and the formed candidate process knowledge set is shown in Table 3.

Table 3 Candidate process knowledge sets obtained based on process intent model

The process information with the highest reliability is pushed by evaluating the process items in the candidate process knowledge set based on the reliability calculation value, and the reliability is S _con <P _type , P _meth-turning >=0.76, S _con <P _{meth- turning} , P _{mach_CK5120} >=0.75, S _con <P _mach , P _clamp-platen >=0.67, and the final pushed process knowledge is roughing-turning-CK5120-platen-YT6/YG8-aqueous.

Claims (10)

In a method of pushing process knowledge based on processing characteristics performed based on an intelligent process design system of machined parts based on a computer network,
following steps,
Step (1): Build a process intention model PPIM based on the processing characteristic vector expression method, where the expression of PPIM is PPIM={PPB, PPG, API}, where PPB is the process design background, and PPG is the process design target and the API is process accessory information;
Step (2): The method for calculating the similarity in the process intention model constructed in the step (1) includes calculating the similarity of parts basic information, calculating the similarity of processing characteristics, and calculating the similarity of quality information;
Step (3): constructing a process knowledge stratified representation model, wherein the process knowledge stratified representation model includes three layers: a basic process information layer, a processing information layer, and a quality information layer;
Step (4): The process knowledge evaluation method based on the reliability calculation includes: obtaining an optimal process knowledge element by using an adjacent process attribute reliability calculation value;
Step (5): Implementing accurate push of process knowledge based on the part process intention model obtained in step (1) and the reliability value in step (4);
Here, the processing characteristic similarity calculation includes three types: vector matching calculation, matrix matching calculation, and attribute value matching calculation,
The vector matching degree and matrix matching degree are obtained by converting the N-th matrix into an N-dimensional vector, and then calculating the vector matching degree between the vector p and the vector q as a relational expression defined by the p vector p _i and the q vector q _i of the i-th element. Calculate and convert the calculated vector matching degree to matrix matching degree to implement matrix matching degree calculation,
Calculation of attribute value matching between a1 attribute value and a2 attribute value is implemented by a relational expression determined by a ₁ , a ₂ attribute values a _1j , a _2j of the j-th element and the number of elements n. How to push process knowledge. According to claim 1,
The construction of the process intention model during the step (1) is constructed according to the part design information and processing requirements, where PPB is the process design background including the part type, blank type, material properties and processing type; PPG is a process design goal including machining characteristic type, tool traversing direction, characteristic surface matrix, proximal surface matrix, topological relationship matrix and specific dimensional information; The API is a process knowledge pushing method based on machining characteristics, characterized in that the process accessory information including geometrical precision information and dimensional precision information. According to claim 1,
The three layers of the process knowledge stratified expression model are specifically,
The basic process information layer includes four process elements, namely, part type, blank type, material attribute and processing type;
The machining information layer includes six process elements, namely machining type, machining method, machine tool type, tool type, fixture type and cutting fluid;
The process knowledge pushing method based on processing characteristics, characterized in that the quality information layer includes specific detection requirements. According to claim 1,
Specific steps of the process knowledge evaluation method based on the reliability calculation;
Step (4.1): determining the priority of each element of process knowledge;
Step (4.2): calculating the reliability of the adjacent process knowledge element, and outputting the process knowledge element with the highest reliability;
Step (4.3): calculating the reliability of the next step process knowledge element until all process knowledge elements are obtained; According to claim 1,
The PPB is obtained through the generated basic design information, and the PPG is obtained through identification processing characteristics, specifically,
The machining property type is mainly determined by a predefined property library including hole properties, slot properties, planar properties and protrusion properties;
The tool feed direction is determined on the basis of the normal vector of the machining surface;
The characteristic surface matrix is formed through each characteristic surface type and its adjacent surface properties. The characteristic surface types include planar, cylindrical, chamfered, spherical, and circular surfaces. , which includes a concave edge, a protruding edge, and a tangent edge;
The topological relationship matrix is determined by the interrelationships between feature planes, including parallel, perpendicular, oblique and tangent;
The basic dimension information is determined by the minimum bounding box based on the machining property, and the process knowledge push method based on the machining property, characterized in that it mainly includes length, width and height. According to claim 1,
The relation between the p vector p _i of the i-th element and the q vector q _i can be expressed as follows,

conine(p, q) represents the degree of matching between the vector p and the vector q, i represents the i-th element of the vector, where the matrix match value is implemented by calculating the degree of matching of the vector,
Process knowledge pushing method based on processing characteristics, characterized in that the relational expression determined by the a ₁ attribute value a _1j and a ₂ attribute value a _2j of the j-th element and the number n of elements is expressed as follows.

S _a (a ₁ , a ₂ ) indicates the degree of matching between attribute values a ₁ and a ₂ , n indicates the number of elements included in the attribute value, and j indicates the j-th element among attribute values. 5. The method of claim 4,
In step (4.1), the priority of each element of the process knowledge refers to the priority relationship of the elements included in the machining information layer, and the order of priority is the machining type, machining method, machine tool type, tool type, and fixture type. and a method of pushing process knowledge based on machining characteristics, characterized in that it is a type of cutting fluid. 5. The method of claim 4,
The calculation formula of the reliability of the adjacent process knowledge factor in step (4.2) is as follows,

where S _con <p ₁ , p ₂ > represents the reliability of process knowledge elements p ₁ and p ₂ , Preq (p ₁ ) represents the total quantity of process knowledge elements p ₁ , and Preq(p ₁ ∩p ₂ ) is a process knowledge pushing method based on processing characteristics, characterized in that represents the related total number of process knowledge elements p ₁ and p ₂ . 6. The method of claim 2 or 5,
The machining characteristic type is expressed as a character string, the tool feed direction is expressed as a vector, and the characteristic face set matrix and the topology relation matrix are generated through attribute type assignment values. According to claim 1,
In the above step (5), based on the obtained part process intention model and the reliability value, a specific step of implementing an accurate push of process knowledge;
Step (5.1): Through the process intention model, it is determined whether the process background information matching degree in the retrieved process knowledge meets the requirement, if so, proceed to step (5.2), otherwise, continue until the requirement is met searching;
Step (5.2): judging whether the process target information matching degree in the process knowledge meets the requirement, if so, proceed to step (5.3), otherwise, continue searching until the requirement is met;
Step (5.3): determining whether the matching degree of process accessory information in the process knowledge meets the requirement, if so, proceed to step (5.4), otherwise, continue searching until the requirement is met;
Step (5.4): outputting a list of process information that meets the requirements, then calculating and obtaining a reliability value through the adjacent process attribute reliability, and finally outputting and pushing the optimal process knowledge element; characterized in that Process knowledge push method based on machining characteristics.

Images