Intelligent design method for recirculating ball type steering gear based on knowledge engineering

Abstract

The invention belongs to the technical field of automobile steering gear intelligence, and particularly discloses a knowledge engineering-based recirculating ball type steering gear intelligent design method, which comprises the following steps of S1, constructing a recirculating ball type steering gear design resource library; s2, receiving design demand parameters input by a user, and retrieving similar historical cases from the design resource library based on a T-KNN algorithm retrieval strategy; s3, parametric modeling and intelligent assembly are conducted on all parts of the steering gear; and S4, based on the engineering drawing knowledge ontology and a preset template, automatically generating an engineering drawing comprising a view, size marks and technical requirements. According to the method, parametric modeling is carried out through the design resource library, automatic generation of an engineering drawing is facilitated, the design period is shortened, rapid retrieval of similar cases is achieved through the T-KNN algorithm, the knowledge reuse rate is increased, dependence on qualification engineers is reduced, and product development is accelerated.

Description

Technical Field

[0001] This invention belongs to the field of intelligent technology of automotive steering systems, and specifically relates to an intelligent design method for a recirculating ball steering system based on knowledge engineering. Background Technology

[0002] As a core component of the automotive steering system, the recirculating ball steering gear is the execution carrier that connects the driver's operating intentions with the vehicle's steering actions. The recirculating ball steering gear assembly consists of key components such as screws, nuts, steel balls, steering gears, and rocker arm shafts. During the design process, the model is frequently adjusted due to changes in various performance parameters.

[0003] For example, when adjusting the diameter of the steering gear screw, the parameters of related components need to be manually modified, which can easily lead to parameter mismatches and interference problems in subsequent models. In the current modeling process of steering gear assemblies, the establishment of geometric models relies on designers manually creating component models one by one, especially the return ball component in recirculating ball steering gears. Its model establishment depends on the return ball curve derived from complex formulas, resulting in low knowledge reuse rate and difficulty in meeting the needs of rapid model establishment. This leads to low design and modeling efficiency and seriously affects the product development cycle.

[0004] Therefore, it is necessary to invent a knowledge engineering-based intelligent design method for recirculating ball steering gears to solve the above problems. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a knowledge engineering-based intelligent design method for recirculating ball steering systems, thereby resolving the issues raised in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a knowledge engineering-based intelligent design method for a recirculating ball steering gear, comprising the following steps: S1. Construct a design resource library for recirculating ball steering gears; S2. Receive design requirement parameters input by the user, and retrieve similar historical cases from the design resource library based on the T-KNN algorithm retrieval strategy; S3. Perform parametric modeling and intelligent assembly of each component of the steering gear. S4. Based on the engineering drawing knowledge ontology and preset templates, automatically generate engineering drawings that include views, dimensions, and technical requirements.

[0007] Furthermore, the ontology of the design resource library is set as follows: RBSG_O={OP, OA, OD, OC}; where OP represents the part knowledge ontology, OA represents the assembly knowledge ontology, OD represents the engineering drawing knowledge ontology, and OC represents the case knowledge ontology.

[0008] Furthermore, the part knowledge ontology is set as: OP={Model_p, Fun_p, Inf_p, Ass_p}; where Model_p includes: a fully constrained 3D model, a set of feature parameters, and driving equations; Fun_p includes: part function type and geometric type; Inf_p includes: storage part number, file path, and design standards; Ass_p includes: assembly features, constraint types such as coincident constraints and distance constraints, and assembly type.

[0009] Furthermore, the assembly knowledge ontology is set as: OA={AF_a, AR_a, AT_a}; where AF_a includes: assembly features in the recirculating ball steering gear assembly knowledge, including reference point elements, reference line elements, and reference surface elements; AR_a includes: assembly relationships in the recirculating ball assembly knowledge; AT_a includes: assembly types in the recirculating ball steering gear assembly knowledge, and the assembly rules are stored in triples (AF_a, AR_a, AT_a).

[0010] Furthermore, the engineering drawing knowledge ontology is set as follows: OD={View_d, Dim_d, Skills_d, Template_d}; where View_d includes: view type and layout coordinates; Dim_d includes: a set of dimensions and datums, clearly describing all dimensions required for manufacturing and inspecting parts, meeting the needs of assembly drawings, part drawings, and assembly, inspection, installation, and use; Skills_d includes: a set of technical requirements, such as geometric tolerances, surface roughness, material requirements, and heat treatment; Template_d includes: engineering drawing template information, drawing size, title block template, and layer specifications.

[0011] Furthermore, the case knowledge ontology is set as: OC={Concept_c, Property_c, Relation_c}; where Concept_c represents the set of case concepts in the case ontology of the recirculating ball steering gear, divides the case hierarchy, and represents the case design features; Property_c includes: case attributes; Relation_c includes: relationships between cases, such as composition relationships, representing the position of a part in the whole; and subordinate relationships, where one concept corresponds to another concept attribute.

[0012] Furthermore, the implementation process in step S2 is as follows: 1. Define case templates based on design requirements, and use these templates to filter matching case sets in the design resource library, narrowing the search scope to 15%-20% of the original design resource library; Second, weights are calculated for attributes, and an AHP (Analytic Hierarchy Process) is used to construct a judgment matrix. Consistency checks are performed to ensure the rationality of subjective weight allocation. Objective weights are then calculated based on similarity deviations. The formula is: ; in, This includes: the similarity between the new requirement and the j-th attribute of the i-th case; The integration method is selected based on the attribute importance α and the numerical difference β. For example, when α > 0.6, "additive integration" is used, and the combined weights are: ; III. Different similarity calculation methods are adopted for different types of attributes. The calculation methods are as follows: For numerical attributes, the similarity calculation formula is as follows: ; Where U is the upper limit of the attribute value and L is the lower limit of the attribute value; For symbolic attributes, the similarity calculation formula is as follows: ; The overall similarity between the new requirement case and the i-th case in the design resource library is calculated by weighted summation, using the following formula: ; in, Let be the similarity of the j-th attribute. Let be the combined weight of the j-th attribute.

[0013] IV. Select cases with a similarity of ≥0.6 and output the case parameters, 3D model path and design document.

[0014] Furthermore, the implementation process in step S3 is as follows: calling the component template library in the design resource library, the engineer inputs the design parameters, and the system automatically drives the model update to achieve parametric modeling; adopting knowledge-driven model intelligent assembly, the system reads the assembly knowledge triplet.

[0015] Furthermore, the parametric modeling implementation process is as follows: Select the part that needs to be 3D modeled in the human-computer interaction interface, determine whether it is a standard part, if it is a standard part, read the standard part model, obtain the template of the corresponding standard part in the corresponding design resource library, and generate a 3D model of the selected model; if it is a non-standard part, obtain the list of key parameters of the current part, input the key parameters and determine whether the parameters are reasonable, if reasonable, generate a 3D model according to the dimensions, and can be modified according to the key dimensions to complete the model establishment.

[0016] Furthermore, the implementation process in step S4 is as follows: match the pre-set "view template", and quickly generate the main view, left view and axial sectional view based on the view template; automatically annotate the dimensions, geometric tolerances and surface roughness of the parts in the view; and automatically fill the technical requirements into the title bar of the view.

[0017] The technical effects and advantages of this invention are as follows: This invention uses a design resource library for parametric modeling, which facilitates the automatic generation of engineering drawings, reduces the design cycle, and utilizes the T-KNN algorithm to achieve rapid retrieval of similar cases, improves knowledge reuse, reduces reliance on senior engineers, and accelerates product development. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart of the intelligent design method for a recirculating ball steering gear based on knowledge engineering, according to an embodiment of the present invention. Figure 2 This is the reuse process of the recirculating ball steering gear design resource library in this embodiment of the invention; Figure 3 This is a flowchart of the intelligent design and parametric modeling of the recirculating ball steering gear according to an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments.

[0020] This invention provides a knowledge engineering-based intelligent design method for recirculating ball steering systems. Figure 1 The process includes the following steps: S1. Construct a design resource library for a recirculating ball steering gear, and use ontology modeling methods to formally represent and store part knowledge (OP), assembly knowledge (OA), engineering drawing knowledge (OD), and case knowledge (OC).

[0021] The ontology of the design resource library is set as: RBSG_O={OP, OA, OD, OC}.

[0022] The part knowledge ontology is set as follows: OP={Model_p, Fun_p, Inf_p, Ass_p}.

[0023] Model_p includes: a fully constrained 3D model, a set of feature parameters, and driving equations. For example, the feature parameter set includes key parameters such as sector gear module, screw pitch, steel ball diameter, and housing wall thickness. The driving equation is set as the formula for calculating the gear center distance. .

[0024] Fun_p includes part function type and geometry type, providing a classification basis for subsequent modeling and assembly. For example, part function types include: transmission components, seals, and support components, and geometry types include: cylindrical gears, shafts, etc.

[0025] Inf_p includes: the storage part number, file path, and design standard. For example, the storage part number can be set to "LF26-520-25".

[0026] Ass_p includes: assembly features, constraint types such as coincident constraints and distance constraints, and assembly types. For example, assembly features include: mating surfaces and locating holes; assembly types include: face-to-face mating and key-and-keyway mating.

[0027] The assembly knowledge ontology is set as follows: OA = {AF_a, AR_a, AT_a}.

[0028] AF_a includes: assembly features in the assembly knowledge of recirculating ball steering gear, including reference point elements, reference line elements, and reference surface elements. For example, the reference point element is set as the center datum point of the part, the reference line element is set as the axis of the shaft part, and the reference surface element is set as the positioning end face and mating surface of the part.

[0029] AR_a includes: assembly relationships of cyclic sphere assembly knowledge. For example, assembly relationships include: coincidence constraints, distance constraints, and angle constraints.

[0030] AT_a includes the assembly type of the recirculating ball steering gear assembly knowledge, storing assembly rules as a triple (AF_a, AR_a, AT_a). For example, assembly types include: gear fit and bolt connection fit.

[0031] The engineering drawing knowledge ontology is set as follows: OD={View_d, Dim_d, Skills_d, Template_d}.

[0032] View_d includes: view type and layout coordinates. For example, view types include: front view, projected view, and sectional view; layout coordinates can be set as follows: in A3 paper, the center coordinates of the front view are (200, 300), and the distance between the projected view and the front view is 50mm, to adapt to the view expression requirements of different types of parts.

[0033] Dim_d includes: a set of dimensions and datums, clearly describing all the dimensions required for manufacturing and inspecting the part, meeting the needs of assembly drawings, part drawings, and assembly, inspection, installation, and use. For example, the set of dimensions includes: defining and locating dimensions; the datum is set as the design reference plane.

[0034] Skills_d includes a set of technical requirements, such as geometric tolerances, surface roughness, material requirements, and heat treatment. For example, the surface roughness is set to Ra1.6, and the material requirement is set to 45 steel or HT200.

[0035] Template_d includes: drawing template information, drawing size, title block template, and layer specifications. For example, the drawing size can be set to: A0, A1, A2, A3, and A4, etc.; the layer specifications include: outline layer and annotation layer.

[0036] The case knowledge ontology is set as: OC={Concept_c, Property_c, Relation_c}.

[0037] Concept_c represents the set of case concepts in the case body of the recirculating ball steering system, dividing the case hierarchy and representing the case design features.

[0038] Property_c includes: Case attributes. For example, front axle load 3500kg, gear ratio 18:1, screw diameter (30mm, 35mm), operating temperature range (-40℃ to 80℃), etc.

[0039] Relation_c includes: relationships between cases, such as composition relationships, which represent the position of a part in the whole; and subordinate relationships, where one concept corresponds to another concept attribute.

[0040] S2. Receive design requirement parameters input by the user and retrieve similar historical cases from the design resource library based on the T-KNN (Template-K Nearest Neighbor) algorithm retrieval strategy. The T-KNN algorithm is used to retrieve similar historical cases from the design resource library, and combined with AHP—Similarity Deviation Combination Weights—to achieve efficient matching of similar cases and knowledge reuse processes, such as... Figure 2 As shown, the implementation process is as follows: I. Define case templates based on design requirements, and use these templates to filter matching case sets in the design resource library, narrowing the search scope to 15%-20% of the original design resource library, thus achieving initial simplification of the search scope. For example, design requirements include parameters such as front axle load and transmission ratio; the case template is based on the "heavy-duty vehicle recirculating ball steering gear template".

[0041] Second, weights are calculated for attributes, and an AHP (Analytic Hierarchy Process) is used to construct a judgment matrix. Consistency checks are performed to ensure the rationality of subjective weight allocation. Objective weights are then calculated based on similarity deviations. The formula is: ; in, This includes the similarity between the new requirement and the j-th attribute of the i-th case.

[0042] For example, the weight of "transmission ratio" in the judgment matrix is ​​set to 0.2, the weight of "center distance" is set to 0.1, and the consistency ratio CR < 0.1.

[0043] The integration method is selected based on the attribute importance α and the numerical difference β. For example, when α > 0.6, "additive integration" is used, and the combined weights are: ; III. Different similarity calculation methods are adopted for different types of attributes. The calculation methods are as follows: For numerical attributes, the similarity calculation formula is as follows: ; Where U is the upper limit of the attribute value and L is the lower limit of the attribute value; for example, the numeric attribute is set to "center distance 20mm".

[0044] For symbolic attributes, the similarity calculation formula is as follows: For example, the symbolic attribute is set to "rotation direction".

[0045] The overall similarity between the new requirement case and the i-th case in the design resource library is calculated by weighted summation, using the following formula: ; in, Let be the similarity of the j-th attribute. Let be the combined weight of the j-th attribute.

[0046] 4. Select cases with a similarity of ≥0.6 and output the case parameters, 3D model path, and design document. For example, the two optimal cases with similarity scores of 0.862 and 0.714 can be selected.

[0047] S3. Perform parametric modeling and intelligent assembly of each component of the steering gear, such as... Figure 3 As shown, the implementation process is as follows: I. Based on Siemens UG secondary development, the system calls the component template library in the design resource library. Engineers input design parameters, and the system automatically drives model updates to achieve parametric modeling. The modeling process is as follows: In the human-computer interaction interface, select the part that needs to be 3D modeled. Determine if it is a standard part. If it is, read the standard part model number, retrieve the corresponding standard part template from the design resource library, and generate a 3D model of the selected model. If it is a non-standard part, retrieve the key parameter list of the current part, input the key parameters, and determine if the parameters are reasonable. If not, retrieve the input key parameters again and determine if they are reasonable. If reasonable, generate a 3D model based on the dimensions, and perform personalized optimization based on the key dimensions. Determine if the preset 3D model is satisfactory. If satisfactory, temporarily save the part's mating dimensions. If not satisfactory, retrieve the key parameter list of the current part again.

[0048] Second, the system adopts a knowledge-driven model for intelligent assembly, and reads assembly knowledge triples.

[0049] Taking the "side cover - O-ring" assembly as an example, in its assembly knowledge OA, AF_a represents the fit between the side cover sealing surface and the outer circle of the O-ring. The assembly relationship AR_a = CC contact fit, and the assembly type AT_a = FF face-to-face fit. The system calculates the part's position matrix P, and the expression for the position matrix is: ; Where S is the rotation matrix, in the form of D is the translation vector, in the form of The system automatically adds assembly constraints to the parts and checks for interference. Taking the "side cover and retaining ring" as an example, the system adds parallel constraints to the two. If the interference is ≤0.02mm, the interference check is considered passed. After the sub-assembly, i.e. the side cover component, is assembled, the system integrates the sub-assemblies into the final assembly.

[0050] S4. Based on the engineering drawing knowledge ontology and preset templates, automatically generate engineering drawings that include views, dimensions, and technical requirements.

[0051] The system first matches a pre-set engineering drawing template and, using the NX engineering drawing API interface, obtains the name of the current part or assembly and the corresponding engineering drawing template file from the design database. Based on this template, it quickly generates the main view, left view, and axial sectional view. Next, the system automatically annotates the dimensions, geometric tolerances, and surface roughness of the part. Simultaneously, the system automatically fills in the technical requirements into the title block of the drawing, including designer information and drawing date. For example, the system first matches a pre-set "A3 rocker arm shaft part drawing template," quickly generating the main view, left view, and axial sectional view in less than 5 minutes, significantly accelerating view generation efficiency. Then, the system automatically annotates the dimensions, geometric tolerances, and surface roughness of the part, such as accurately annotating the fan tooth tip circle diameter of 128mm, circular runout tolerance of 0.06mm, and surface roughness Ra1.6. At the same time, the system automatically fills in the technical requirements into the title block of the drawing.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it.

Claims

1. A knowledge engineering-based intelligent design method for a recirculating ball steering gear, characterized in that: Includes the following steps: S1. Construct a design resource library for recirculating ball steering gears; S2. Receive design requirement parameters input by the user, and retrieve similar historical cases from the design resource library based on the T-KNN algorithm retrieval strategy; S3. Perform parametric modeling and intelligent assembly of each component of the steering gear. S4. Based on the engineering drawing knowledge ontology and preset templates, automatically generate engineering drawings that include views, dimensions, and technical requirements.

2. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 1, characterized in that: The ontology of the design resource library is set as follows: RBSG_O={OP, OA, OD, OC}; Wherein, OP represents the part knowledge ontology, OA represents the assembly knowledge ontology, OD represents the engineering drawing knowledge ontology, and OC represents the case knowledge ontology.

3. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 2, characterized in that: The part knowledge ontology is set as follows: OP={Model_p, Fun_p, Inf_p, Ass_p}; Model_p includes: a fully constrained 3D model, a set of feature parameters, and driving equations; Fun_p includes: part function type and geometry type; Inf_p includes: storage part number, file path, and design standard; Ass_p includes: assembly features, constraint types such as coincident constraints and distance constraints, and assembly types.

4. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 2, characterized in that: The assembly knowledge ontology is set as follows: OA={AF_a, AR_a, AT_a}; Among them, AF_a includes: assembly features in the assembly knowledge of recirculating ball steering gear, including reference point elements, reference line elements, and reference surface elements; AR_a includes: the assembly relationships of the knowledge of cyclic ball assembly; AT_a includes: assembly type of recirculating ball steering gear assembly knowledge, which stores assembly rules as a triple (AF_a, AR_a, AT_a).

5. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 2, characterized in that: The engineering drawing knowledge ontology is set as follows: OD={View_d, Dim_d, Skills_d, Template_d}; View_d includes: view type and layout coordinates; Dim_d includes: a set of dimensions and datums, clearly describing all the dimensions required for manufacturing and inspecting parts, meeting the needs of assembly drawings, part drawings, and assembly, inspection, installation, and use; Skills_d includes: a set of technical requirements, such as geometric tolerances, surface roughness, material requirements, and heat treatment; Template_d includes: engineering drawing template information, drawing size, title block template, and layer specifications.

6. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 2, characterized in that: The case knowledge ontology is set as: OC={Concept_c, Property_c, Relation_c}; Where Concept_c represents the set of case concepts in the case body of the recirculating ball steering gear, which divides the case hierarchy and represents the case design features; Property_c includes: Case Properties; Relation_c includes: relationships between cases, such as composition relationships, which represent the position of a part in the whole; and subordinate relationships, where one concept corresponds to another concept attribute.

7. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 1, characterized in that: The implementation process in step S2 is as follows:

1. Define case templates based on design requirements, and use these templates to filter matching case sets in the design resource library, narrowing the search scope to 15%-20% of the original design resource library; Second, weights are calculated for attributes, and an AHP (Analytic Hierarchy Process) is used to construct a judgment matrix. Consistency checks are performed to ensure the rationality of subjective weight allocation. Objective weights are then calculated based on similarity deviations. The formula is: ； in, This includes: the similarity between the new requirement and the j-th attribute of the i-th case; The integration method is selected based on the attribute importance α and the numerical difference β. For example, when α > 0.6, "additive integration" is used, and the combined weights are: ; III. Different similarity calculation methods are adopted for different types of attributes. The calculation methods are as follows: For numerical attributes, the similarity calculation formula is as follows: ； Where U is the upper limit of the attribute value and L is the lower limit of the attribute value; For symbolic attributes, the similarity calculation formula is as follows: ； The overall similarity between the new requirement case and the i-th case in the design resource library is calculated by weighted summation, using the following formula: ； in, Let be the similarity of the j-th attribute. The combined weight of the j-th attribute; IV. Select cases with a similarity of ≥0.6 and output the case parameters, 3D model path and design document.

8. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 4, characterized in that: The implementation process in step S3 is as follows: The system calls upon the component template library in the design resource library, and engineers input design parameters. The system then automatically drives model updates to achieve parametric modeling. The system employs a knowledge-driven model for intelligent assembly, reading assembly knowledge triples.

9. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 8, characterized in that: The parametric modeling implementation process is as follows: In the human-computer interaction interface, select the part that needs to be 3D modeled, determine whether it is a standard part, if it is a standard part, read the standard part model, obtain the template of the corresponding standard part from the corresponding design resource library, and generate the 3D model of the selected model. If it is a non-standard part, obtain the list of key parameters of the current part, input the key parameters and judge whether the parameters are reasonable. If they are reasonable, generate a 3D model based on the dimensions, and modify them according to the key dimensions to complete the model building.

10. The intelligent design method for a recirculating ball steering gear based on knowledge engineering according to claim 1, characterized in that: The implementation process in step S4 is as follows: Match the pre-set "view template" and quickly generate the main view, left view, and axial section view based on the view template; Automatically annotate the dimensions, geometric tolerances, and surface roughness of the parts in the view; Automatically populate the technical requirements into the view's title bar.

Images