A system and method for automatically correcting homework

Abstract

The present invention discloses a system and method for automatic homework correction, and relates to the field of automated intelligent technology. The method can parse and extract geometric data and geometric feature sets in real time by importing mechanical parts design drawing homework into CAD software, and combining CAD parsing tools, and then integrate with automatic correction application programs, and can automatically extract without manual intervention, and can automatically perform dimensional error analysis, geometric structure rationality analysis, material selection verification and geometric symmetry analysis. Compared with the traditional manual correction method, the automatic correction method not only saves time, but also significantly improves efficiency, especially when facing a large number of homework, it can quickly complete the correction task, reduce the workload of teachers, and avoid the occurrence of human errors. The automated correction method ensures the accuracy and consistency of the correction process, and is particularly suitable for large-class teaching or the processing of complex design drawings.

Description

Technical Field

[0001] The present invention relates to the field of automated intelligent technology, and in particular to an automatic homework grading system and method. Background Art

[0002] The automatic grading method for homework belongs to an important innovative direction in the field of modern educational technology and intelligent system development. Specifically, it focuses on engineering education based on computer-aided design (CAD), especially the automated evaluation and feedback of mechanical design drawings. Mechanical part design drawings are one of the important forms of homework in engineering teaching. Drawings drawn and designed by CAD software must not only meet strict geometric requirements, but also be accurate in size, structural rationality, and material selection.

[0003] At present, due to the limitations of manual grading, many design drawings submitted by students often have some potential problems that cannot be corrected in time. This not only affects students' understanding of the design drawings, but may also cause them to repeat the same mistakes in subsequent designs. In addition, due to time and energy constraints, teachers are unable to provide personalized guidance to each student, resulting in students lacking clear direction when revising their homework. Ultimately, this situation may affect the improvement of students' design capabilities, especially when it comes to complex mechanical parts. The accumulation of errors may lead to unreasonable designs and even affect future engineering practices. The lack of automatic homework grading methods not only affects teaching efficiency, but also hinders the improvement of students' learning outcomes. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a system and method for automatically correcting homework, which solves the problems mentioned in the background technology.

[0005] A method for automatically correcting homework, in order to achieve the above purpose, the present invention is implemented by the following technical scheme: comprising the following steps:

[0006] S1. Import the submitted mechanical part design drawing job into the CAD software, extract the geometric data of the mechanical part design drawing job, and parse the mechanical part design drawing job through the CAD parsing tool to obtain the geometric feature set, and integrate the CAD software with the automatic correction application to extract the obtained geometric data and geometric feature set in real time;

[0007] S2. Through the automatic correction application, based on the acquired geometric data and geometric feature set, dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis are performed, and the dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry are output;

[0008] S3, comprehensively calculate the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, and output the comprehensive score Tscore of the job, and make a preliminary comprehensive evaluation based on the preset job qualification threshold T and the obtained comprehensive score Tscore, and correct the qualification of the currently submitted mechanical part design drawing job;

[0009] S4. After preliminary comprehensive evaluation, the identifier TSid is generated. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry;

[0010] S5. After modifying the drawings according to the rectification suggestions, submit the mechanical parts design drawing job for the second time, verify the uniqueness of the drawings through the identifier TSid, and then execute S1 to S3, output the comprehensive score Tscore2 of the second job, calculate the output total score difference △Tscore, and generate a design difference report.

[0011] Preferably, S1 includes S11, S12 and S13;

[0012] S11, importing the submitted mechanical part design drawing work into the CAD software, then identifying the geometric data in the mechanical part design drawing work through the Measure tool, and exporting the geometric data into a table file;

[0013] The geometric data include the design size dstudent, the hole diameter Dhole, the shaft diameter Dshaft, the length of the left side of the symmetry axis Lleft and the length of the right side of the symmetry axis Lright;

[0014] S12, automatically analyzing the mechanical parts design drawing operation through the CAD analysis function to obtain the geometric feature set;

[0015] The geometric feature set includes connection tolerance TOL, geometric structure characteristic coefficient Gshape, and material property characteristic parameter Pmaterial;

[0016] The connection tolerance TOL is obtained by extracting the hole diameter Dhole and shaft diameter Dshaft from the table file corresponding to the geometric data for tolerance calculation output. The specific algorithm formula is: ;

[0017] The geometric structure feature coefficient Gshape is directly extracted from the mechanical part design drawing by CAD parsing tools. The geometric structure features include but are not limited to holes, slots, bosses and complex surfaces. Then, for each geometric structure feature i, the sub-feature k is obtained according to the functional requirements of the design. The sub-features include but are not limited to the aperture, wall thickness and curvature. The geometric structure feature coefficient Gshape is calculated. The specific algorithm formula is: ;In the formula, Gshape i represents the i-th geometric structure characteristic coefficient, n i represents the total number of geometric structure features of the i-th geometry, f represents the evaluation function, and Xshape i，k Represents the kth sub-feature parameter value of the i-th geometric structure feature, Crequire i，k The standard value of the kth sub-feature parameter of the i-th geometric structure feature is initially set by the teacher according to the homework situation;

[0018] The material property characteristic parameter Pmaterial is directly extracted and matched with the material database through the CAD parsing tool. Once the material type is determined, the system extracts the corresponding material performance parameters from the material database, and the material properties in the material database are directly used as preset values;

[0019] S13. By building an automatic correction application and setting an API application interface to integrate with the CAD software, after the mechanical part design drawing is imported into the CAD software and the geometric data and geometric feature set are obtained, the geometric data and geometric feature set are automatically extracted through the API application interface.

[0020] Preferably, S2 includes S21, S22, S23 and S24;

[0021] S21, through the automatic correction application, based on the acquired geometric data and geometric feature set, extract the connection tolerance TOL and the design dimension dstudent for dimensional error analysis, output the dimensional accuracy score Sdim, and perform dimensional error analysis;

[0022] The dimensional accuracy score Sdim is calculated using the following algorithm formula:

[0023] ;

[0024] In the formula, C represents the total number of design dimensions, represents the standardized error, dstandard represents the standard design size, represents the standardized error of the c-th design dimension.

[0025] Preferably, S22, by extracting the i-th geometric structure feature coefficient Gshape in the geometric feature set i , through the geometric shape matching algorithm, calculate and output the geometric structure rationality score Fvalidity, and analyze the geometric structure rationality;

[0026] The geometric structure rationality score Fvalidity is calculated and output by the following geometric shape matching algorithm;

[0027] ;

[0028] Where n represents the total number of geometric structure features, wi represents the weight value of the i-th geometric structure feature, and its specific value is set by the user according to the design requirements of mechanical parts.

[0029] Preferably, S23, by extracting the material property characteristic parameter Pmaterial from the geometric feature set, calculating and outputting the material selection rationality score Mmatch through the material matching evaluation formula, analyzing the matching degree between the material performance in the mechanical part design drawing operation and the design requirements, and verifying the selection of the annotated material information for the mechanical part design drawing operation;

[0030] The material selection rationality score Mmatch is calculated and output by the following material matching evaluation formula;

[0031] ;

[0032] Where m represents the number of material properties, which means the number of material properties that need to be evaluated, such as tensile strength, density, and hardness, etc. j Represents the characteristic parameter of the j-th material property, Prequired j represents the characteristic parameters of the material properties required for the j-th design, and Mj represents the weight value corresponding to the j-th material property.

[0033] Preferably, S24, by extracting the left side length Lleft and the right side length Lright of the symmetry axis in the geometric data, calculating and outputting the part geometric symmetry score Ssymmctry, and performing a symmetry rationality analysis on the geometric structure;

[0034] The part geometric symmetry score Ssymmctry is calculated and output by the following algorithm formula;

[0035] ;

[0036] Where Lleft i represents the left symmetric length of the i-th geometric structure feature, Lright irepresents the right symmetric length of the i-th geometric structure feature, and n represents the total number of geometric structure features.

[0037] Preferably, S3 includes S31 and S32;

[0038] S31, based on the obtained dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, perform comprehensive calculation and output the comprehensive score Tscore of the job;

[0039] The comprehensive score Tscore of the assignment is calculated using the following algorithm formula;

[0040] ;

[0041] Where a1, a2, a3 and a4 represent the preset weight values ​​of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry, respectively, and a1+a2+a3+a4=1, and the specific values ​​are set by the user;

[0042] S32, based on the user-set job qualification threshold T and the obtained job comprehensive score Tscore, a preliminary comprehensive evaluation is performed to analyze the qualification of the mechanical parts design drawing job submitted for the first time. The specific evaluation contents are as follows;

[0043] When the comprehensive score of the job Tscore ≥ the job qualification threshold T, it means that the mechanical parts design drawing job is qualified;

[0044] When the comprehensive score Tscore of the job is less than the job qualification threshold T, it means that the mechanical parts design drawing job is unqualified.

[0045] Preferably, S4 includes S41 and S42;

[0046] S41. After the preliminary comprehensive evaluation, each mechanical part design drawing is assigned a unique identifier TSid. After the identifier TSid is set, it is transferred to the automatic correction application database. In order to facilitate subsequent query, tracking and modification, the system will generate a unique identifier TSid for each design. This identifier not only corresponds to the score and feedback report of the design, but also is associated with the detailed information of each module of the design drawing;

[0047] The identifier TSid is generated as follows: , where TimeStamp represents the timestamp, CourseID represents the course ID, and StudentID represents the student ID;

[0048] S42. When the preliminary comprehensive assessment shows that the mechanical parts design drawing is unqualified, the initial set size requirement index Cc, geometric structure rationality requirement index Hl, material requirement index Cl and symmetry requirement index Dc are automatically compared with the size accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry based on the specific values ​​set by the teacher according to the design requirements, and the output results are obtained and modification suggestions are generated. The specific modification suggestions are as follows;

[0049] When the dimensional accuracy score Sdim≤dimensional requirement index Cc, it meets the requirements and no prompt is needed;

[0050] When the dimensional accuracy score Sdim>dimensional requirement index Cc, it does not meet the requirements and is marked as needing adjustment. The modification suggestion prompts to optimize the dimensional accuracy.

[0051] When the geometric structure rationality score Fvalidity ≥ geometric structure rationality requirement index Hl, the requirement is met and no prompt is needed;

[0052] When the geometric structure rationality score Fvalidity is less than the geometric structure rationality requirement index Hl, it does not meet the requirements and is marked as needing adjustment. The rationality of optimizing the geometric structure is prompted in the modification suggestions;

[0053] When the material selection rationality score Mmatch≥material requirement index Cl, it meets the requirements and no prompt is needed;

[0054] When the material selection rationality score Mmatch is less than the material requirement index Cl, it does not meet the requirements and is marked as requiring adjustment. In the modification suggestions, it is suggested to optimize the rationality of material selection;

[0055] When the part geometric symmetry score Ssymmctry = symmetry requirement index Dc, the part symmetry meets the requirements and no prompt is needed;

[0056] When the part geometric symmetry score Ssymmctry is less than the symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometric structure symmetry;

[0057] When the part geometry symmetry score Ssymmctry>symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometry symmetry.

[0058] Preferably, S5 includes S51 and S52;

[0059] S51. After modifying the mechanical part design drawing job according to the rectification suggestion, submit the mechanical part design drawing job for the second time, and verify the uniqueness of the mechanical part design drawing job by inputting the identifier TSid into the automatic correction application;

[0060] S52, after verifying the uniqueness, execute S1 to S3 for the second time, output the second job comprehensive score Tscore2, and perform difference calculation with the job comprehensive score Tscore to output the total score difference △Tscore, and generate a design difference report based on the total score difference △Tscore output result;

[0061] The total score difference △Tscore is calculated and output by the following algorithm formula;

[0062] ;

[0063] The design difference report is generated by the following content:

[0064] When the total score difference △Tscore>0, it means that there is improvement in the second submission of mechanical parts design drawing work, and no difference report is generated at this time;

[0065] When the total score difference △Tscore≤0, a difference report is generated, indicating that the modification of the mechanical part design drawing assignment submitted for the second time is abnormal. At the same time, the instructor's name and contact information are generated in the report, prompting you to find a teacher for guidance.

[0066] An automatic homework grading system includes a homework submission analysis module, a grading analysis module, a comprehensive grading module, a feedback and optimization suggestion module, and a secondary evaluation difference analysis module;

[0067] The job submission analysis module imports the submitted mechanical part design drawing job into the CAD software, extracts the geometric data of the mechanical part design drawing job, and parses the mechanical part design drawing job through the CAD parsing tool to obtain a geometric feature set, and integrates the CAD software with the automatic correction application to extract the acquired geometric data and geometric feature set in real time;

[0068] The correction analysis module performs dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis based on the acquired geometric data and geometric feature set through an automatic correction application, and outputs a dimensional accuracy score Sdim, a geometric structure rationality score Fvalidity, a material selection rationality score Mmatch and a part geometric symmetry score Ssymmctry;

[0069] The comprehensive correction module comprehensively calculates the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, outputs the comprehensive score Tscore of the job, and performs a preliminary comprehensive evaluation on the preset job qualification threshold T and the obtained comprehensive score Tscore of the job, and corrects the qualification of the currently submitted mechanical part design drawing job;

[0070] The feedback and optimization suggestion module generates an identifier TSid after a preliminary comprehensive evaluation. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry.

[0071] After the secondary evaluation difference analysis module modifies the drawing through rectification suggestions, it submits the mechanical part design drawing job for the second time, verifies the uniqueness of the drawing through the identifier TSid, and then executes S1 to S3, outputs the second job comprehensive score Tscore2, calculates the output total score difference △Tscore, and generates a design difference report.

[0072] The present invention provides a system and method for automatically correcting homework, which has the following beneficial effects:

[0073] (1) This method imports mechanical parts design drawings into CAD software and combines them with CAD parsing tools. The automatic grading system can extract geometric data and geometric feature sets in real time. The system can automatically perform dimensional error analysis, geometric structure rationality analysis, material selection verification, and geometric symmetry analysis without human intervention. Compared with the traditional manual grading method, the automatic grading method not only saves time, but also significantly improves efficiency. Especially when faced with a large amount of homework, it can quickly complete the grading task, reduce the workload of teachers, and avoid the occurrence of human errors. The automated grading method ensures the accuracy and consistency of the grading process, which is particularly suitable for large-class teaching or the processing of complex design drawings.

[0074] (2) This method can comprehensively cover the key parameters in the design drawings based on the calculation of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch, and the geometric symmetry score Ssymmetry. This multi-dimensional automatic evaluation mechanism can not only give a comprehensive score Tscore for the assignment, but also provide detailed feedback on unqualified designs. The system automatically triggers the feedback mechanism and generates targeted modification suggestions based on the analysis results of each scoring item, such as providing optimization directions in terms of dimensional accuracy, geometric structure rationality, material selection rationality, and symmetry. Through these personalized feedbacks, students can clearly know where improvements need to be made, so as to make targeted design modifications.

[0075] (3) This method also supports students to submit their assignments twice. After each submission, the system generates a unique identifier TSid to ensure the uniqueness of each design drawing assignment. Through the identifier, the system can compare the score difference between the new and old design versions during the second submission and calculate the total score difference ΔTscore. When the second assignment is improved, the system will generate a design difference report to help students understand the effectiveness of the modification. When the improvement effect is not obvious, the system will also prompt students to further optimize and provide the instructor's information for help. Through such a tracking and feedback mechanism, students can gradually optimize the design and improve their understanding of the design requirements. At the same time, teachers can also provide targeted guidance based on the students' submission situation to improve the overall teaching effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 A schematic diagram of the steps of an automatic homework correction method of the present invention;

[0077] Figure 2 The present invention is a flowchart of an automatic homework correction system. DETAILED DESCRIPTION

[0078] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0079] Example 1

[0080] See also Figure 1 The present invention provides a method for automatically correcting homework. To achieve the above purpose, the present invention is implemented through the following technical scheme: comprising the following steps:

[0081] S1. Import the submitted mechanical part design drawing job into the CAD software, extract the geometric data of the mechanical part design drawing job, and parse the mechanical part design drawing job through the CAD parsing tool to obtain the geometric feature set, and integrate the CAD software with the automatic correction application to extract the obtained geometric data and geometric feature set in real time;

[0082] S2. Through the automatic correction application, based on the acquired geometric data and geometric feature set, dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis are performed, and the dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry are output;

[0083] S3, comprehensively calculate the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, and output the comprehensive score Tscore of the job, and make a preliminary comprehensive evaluation based on the preset job qualification threshold T and the obtained comprehensive score Tscore, and correct the qualification of the currently submitted mechanical part design drawing job;

[0084] S4. After preliminary comprehensive evaluation, the identifier TSid is generated. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry;

[0085] S5. After modifying the drawings according to the rectification suggestions, submit the mechanical parts design drawing job for the second time, verify the uniqueness of the drawings through the identifier TSid, and then execute S1 to S3, output the comprehensive score Tscore2 of the second job, calculate the output total score difference △Tscore, and generate a design difference report.

[0086] In this embodiment, the method can automatically extract geometric data and feature sets by importing mechanical parts design drawings into CAD software through step S1, combined with automatic correction application, so as to realize efficient digital processing of operation data and improve the accuracy and real-time performance of data acquisition. Step S2 generates accurate scores by multi-dimensional analysis of geometric data, including dimensional error, rationality of geometric structure, material selection verification, etc., and provides a comprehensive and accurate evaluation for each design. Step S3 combines various scores, automatically calculates the total score Tscore of the assignment, and compares it with the preset threshold, thereby realizing a fast and reliable qualification evaluation. Step S4 ensures that unqualified assignments can automatically trigger the feedback mechanism, provide specific rectification suggestions, and enable students to accurately identify and modify problems in the design. Step S5 generates a design difference report by supporting secondary submission and score comparison, which helps to track the improvement progress of students and improves the feedback effect and learning efficiency of teaching. Compared with traditional manual correction, this automatic correction method greatly improves the efficiency, accuracy and consistency of evaluation, reduces human errors and teacher workload. Through real-time data processing and multi-dimensional scoring, the system can more accurately identify problems in the design and provide personalized feedback, thereby helping students to make more timely improvements in their designs. This method not only improves the overall efficiency of grading, but also optimizes the student learning process, allowing them to continuously improve their designs and achieve significant improvements in technical means.

[0087] Example 2

[0088] This embodiment is explained in Example 1, please refer to Figure 1 , specifically: S1 includes S11, S12 and S13;

[0089] S11, importing the submitted mechanical part design drawing work into the CAD software, then identifying the geometric data in the mechanical part design drawing work through the Measure tool, and exporting the geometric data into a table file;

[0090] The geometric data include the design size dstudent, the hole diameter Dhole, the shaft diameter Dshaft, the length of the left side of the symmetry axis Lleft and the length of the right side of the symmetry axis Lright;

[0091] S12, automatically analyzing the mechanical parts design drawing operation through the CAD analysis function to obtain the geometric feature set;

[0092] The geometric feature set includes connection tolerance TOL, geometric structure characteristic coefficient Gshape, and material property characteristic parameter Pmaterial;

[0093] The connection tolerance TOL is obtained by extracting the hole diameter Dhole and shaft diameter Dshaft from the table file corresponding to the geometric data for tolerance calculation output. The specific algorithm formula is: ;

[0094] The geometric structure feature coefficient Gshape is directly extracted from the mechanical part design drawing by CAD parsing tools. The geometric structure features include but are not limited to holes, slots, bosses and complex surfaces. Then, for each geometric structure feature i, the sub-feature k is obtained according to the functional requirements of the design. The sub-features include but are not limited to the aperture, wall thickness and curvature. The geometric structure feature coefficient Gshape is calculated. The specific algorithm formula is: ;In the formula, Gshape i represents the i-th geometric structure characteristic coefficient, n i represents the total number of geometric structure features of the i-th geometry, f represents the evaluation function, and Xshape i，k Represents the kth sub-feature parameter value of the i-th geometric structure feature, Crequire i，k The standard value of the kth sub-feature parameter of the i-th geometric structure feature is initially set by the teacher according to the homework situation;

[0095] The material property characteristic parameter Pmaterial is directly extracted and matched with the material database through the CAD parsing tool. Once the material type is determined, the system extracts the corresponding material performance parameters from the material database, and the material properties in the material database are directly used as preset values;

[0096] S13. By building an automatic correction application and setting an API application interface to integrate with the CAD software, after the mechanical part design drawing is imported into the CAD software and the geometric data and geometric feature set are obtained, the geometric data and geometric feature set are automatically extracted through the API application interface.

[0097] In this embodiment, S11 of the method realizes the automatic acquisition of design information by importing the design drawings into the CAD software and automatically extracting the geometric data using the Measure tool, which greatly improves the efficiency and accuracy of data processing. Compared with traditional manual measurement and data entry, the automatic export of geometric data as a table file eliminates the error of manual input and improves the speed of data analysis. S12 further utilizes the analytical function of CAD to extract the geometric feature set, and enables the system to accurately evaluate the rationality of the design by automatically calculating the connection tolerance and the geometric feature coefficient. This not only speeds up the analytical process of complex geometric structures, but also ensures the match between the design and functional requirements. S13 realizes the automation of real-time interaction and correction of data by automatically extracting and processing geometric data and feature sets with the help of API interface through the integration of automatic correction application and CAD. Compared with the traditional manual correction method, this method not only eliminates the inaccuracy in manual operation, but also significantly improves the correction speed and feedback efficiency, further making the correction process in mechanical design teaching intelligent and automated, thereby achieving an overall improvement in effect.

[0098] Example 3

[0099] This embodiment is explained in Example 2. Please refer to Figure 1 , specifically: S2 includes S21, S22, S23 and S24;

[0100] S21, through the automatic correction application, based on the acquired geometric data and geometric feature set, extract the connection tolerance TOL and the design dimension dstudent for dimensional error analysis, output the dimensional accuracy score Sdim, and perform dimensional error analysis;

[0101] The dimensional accuracy score Sdim is calculated using the following algorithm formula:

[0102] ;

[0103] In the formula, C represents the total number of design dimensions, represents the standardized error, dstandard represents the standard design size, represents the standardized error of the c-th design dimension.

[0104] S22, by extracting the i-th geometric structure feature coefficient Gshape from the geometric feature set i , through the geometric shape matching algorithm, calculate and output the geometric structure rationality score Fvalidity, and analyze the geometric structure rationality;

[0105] The geometric structure rationality score Fvalidity is calculated and output by the following geometric shape matching algorithm;

[0106] ;

[0107] Where n represents the total number of geometric structure features, wi represents the weight value of the i-th geometric structure feature, and its specific value is set by the user according to the design requirements of mechanical parts.

[0108] S23, by extracting the material property characteristic parameter Pmaterial from the geometric feature set, using the material matching evaluation formula, calculating and outputting the material selection rationality score Mmatch, analyzing the matching degree between the material performance in the mechanical part design drawing operation and the design requirements, and verifying the selection of the material information for the mechanical part design drawing operation;

[0109] The material selection rationality score Mmatch is calculated and output by the following material matching evaluation formula;

[0110] ;

[0111] Where m represents the number of material properties, which means the number of material properties that need to be evaluated, such as tensile strength, density, and hardness, etc. j Represents the characteristic parameter of the j-th material property, Prequired j represents the characteristic parameters of the material properties required for the j-th design, and Mj represents the weight value corresponding to the j-th material property.

[0112] S24, by extracting the left side length Lleft and the right side length Lright of the symmetry axis in the geometric data, calculating and outputting the part geometric symmetry score Ssymmctry, and performing a symmetry rationality analysis on the geometric structure;

[0113] The part geometric symmetry score Ssymmctry is calculated and output by the following algorithm formula;

[0114] ;

[0115] Where Lleft i represents the left symmetric length of the i-th geometric structure feature, Lright i represents the right symmetric length of the i-th geometric structure feature, and n represents the total number of geometric structure features.

[0116] In this embodiment, S21 of the method extracts geometric data and feature sets, performs dimensional error analysis, automatically calculates and outputs the dimensional accuracy score Sdim, and ensures that the size of the part meets the design standards. Compared with traditional manual measurement and analysis, this step significantly improves the accuracy and speed of dimensional evaluation. S22 calculates the geometric structure rationality score Fvalidity through the extraction and matching algorithm of geometric structure features, effectively evaluates the functional rationality of the part, and ensures that the structural design meets the use requirements. This process overcomes the shortcomings of manual correction that may omit structural function problems. S23 uses the material matching evaluation formula to output the material selection rationality score Mmatch based on the material property characteristic parameter Pmaterial, automatically verifies whether the selected material meets the design requirements, and makes the material evaluation process more efficient and reliable. S24 automatically calculates the part geometric symmetry score Ssymmetry by extracting the lengths on the left and right sides of the symmetry axis to ensure that the symmetry of the part meets the design standards. Compared with existing technical means, this automated correction method significantly improves the accuracy and efficiency of dimension, geometry, material and symmetry assessment, solves the time-consuming and error-prone problems commonly seen in manual correction, and thus improves the overall quality and efficiency of design drawing correction.

[0117] Example 4

[0118] This embodiment is explained in Example 3, please refer to Figure 1 , specifically: S3 includes S31 and S32;

[0119] S31, based on the obtained dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, perform comprehensive calculation and output the comprehensive score Tscore of the job;

[0120] The comprehensive score Tscore of the assignment is calculated using the following algorithm formula;

[0121] ;

[0122] Where a1, a2, a3 and a4 represent the preset weight values ​​of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry, respectively, and a1+a2+a3+a4=1, and the specific values ​​are set by the user;

[0123] S32, based on the user-set job qualification threshold T and the obtained job comprehensive score Tscore, a preliminary comprehensive evaluation is performed to analyze the qualification of the mechanical parts design drawing job submitted for the first time. The specific evaluation contents are as follows;

[0124] When the comprehensive score of the job Tscore ≥ the job qualification threshold T, it means that the mechanical parts design drawing job is qualified;

[0125] When the comprehensive score Tscore of the job is less than the job qualification threshold T, it means that the mechanical parts design drawing job is unqualified.

[0126] In this embodiment, S31 of the method uses a weighted algorithm to automatically calculate the comprehensive score Tscore of the assignment by combining the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the geometric symmetry score Ssymmetry. This process realizes a comprehensive evaluation of the drawing assignment, rationally integrates each scoring dimension, and ensures that the importance of different design elements is fully reflected. At the same time, users can set the weights of each scoring item according to specific needs, making the scoring mechanism more flexible and personalized. S32 quickly obtains the qualification result of the assignment by comparing the calculated comprehensive score Tscore with the preset assignment qualification threshold T. This automated evaluation mechanism greatly shortens the time required for traditional manual correction and improves the accuracy and consistency of judgment. Compared with the current manual scoring, which is highly subjective and error-prone, this method realizes a more efficient and reliable correction process through a precise algorithm, ensuring that the evaluation results of each assignment can be quickly generated with high accuracy and fairness, and improving the overall effect of teaching correction.

[0127] Example 5

[0128] This embodiment is explained in Example 4. Please refer to Figure 1 , specifically: S4 includes S41 and S42;

[0129] S41. After the preliminary comprehensive evaluation, each mechanical part design drawing is assigned a unique identifier TSid. After the identifier TSid is set, it is transferred to the automatic correction application database. In order to facilitate subsequent query, tracking and modification, the system will generate a unique identifier TSid for each design. This identifier not only corresponds to the score and feedback report of the design, but also is associated with the detailed information of each module of the design drawing;

[0130] The identifier TSid is generated as follows: , where TimeStamp represents the timestamp, CourseID represents the course ID, and StudentID represents the student ID;

[0131] S42. When the preliminary comprehensive assessment shows that the mechanical parts design drawing is unqualified, the initial set size requirement index Cc, geometric structure rationality requirement index Hl, material requirement index Cl and symmetry requirement index Dc are automatically compared with the size accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry based on the specific values ​​set by the teacher according to the design requirements, and the output results are obtained and modification suggestions are generated. The specific modification suggestions are as follows;

[0132] When the dimensional accuracy score Sdim≤dimensional requirement index Cc, it meets the requirements and no prompt is needed;

[0133] When the dimensional accuracy score Sdim>dimensional requirement index Cc, it does not meet the requirements and is marked as needing adjustment. The modification suggestion prompts to optimize the dimensional accuracy.

[0134] When the geometric structure rationality score Fvalidity ≥ geometric structure rationality requirement index Hl, the requirement is met and no prompt is needed;

[0135] When the geometric structure rationality score Fvalidity is less than the geometric structure rationality requirement index Hl, it does not meet the requirements and is marked as needing adjustment. The rationality of optimizing the geometric structure is prompted in the modification suggestions;

[0136] When the material selection rationality score Mmatch≥material requirement index Cl, it meets the requirements and no prompt is needed;

[0137] When the material selection rationality score Mmatch is less than the material requirement index Cl, it does not meet the requirements and is marked as requiring adjustment. In the modification suggestions, it is suggested to optimize the rationality of material selection;

[0138] When the part geometric symmetry score Ssymmctry = symmetry requirement index Dc, the part symmetry meets the requirements and no prompt is needed;

[0139] When the part geometric symmetry score Ssymmctry is less than the symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometric structure symmetry;

[0140] When the part geometry symmetry score Ssymmctry>symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometry symmetry.

[0141] In this embodiment, S41 of the method realizes personalized management of each design drawing by generating a unique identifier TSid for each assignment. This identifier can not only be associated with the comprehensive score and feedback report of the design, but also can be associated with all detailed module information of the design drawing to ensure the efficiency of subsequent query, tracking and modification. This process greatly improves the systematization and traceability of homework management, and facilitates teachers and students to track and modify the homework for a long time. S42 automatically generates detailed modification suggestions for unqualified homework based on the evaluation criteria preset by the system. Students can understand the deficiencies in the homework through these specific feedbacks, and make modifications one by one according to the feedback of the scoring items, such as optimizing dimensional accuracy, improving geometric structure, optimizing material selection or adjusting symmetry design.

[0142] Example 6

[0143] This embodiment is explained in Example 5. Please refer to Figure 1 , specifically: S5 includes S51 and S52;

[0144] S51. After modifying the mechanical part design drawing job according to the rectification suggestion, submit the mechanical part design drawing job for the second time, and verify the uniqueness of the mechanical part design drawing job by inputting the identifier TSid into the automatic correction application;

[0145] S52, after verifying the uniqueness, execute S1 to S3 for the second time, output the second job comprehensive score Tscore2, and perform difference calculation with the job comprehensive score Tscore to output the total score difference △Tscore, and generate a design difference report based on the total score difference △Tscore output result;

[0146] The total score difference △Tscore is calculated and output by the following algorithm formula;

[0147] ;

[0148] The design difference report is generated by the following content:

[0149] When the total score difference △Tscore>0, it means that there is improvement in the second submission of mechanical parts design drawing work, and no difference report is generated at this time;

[0150] When the total score difference △Tscore≤0, a difference report is generated, indicating that the modification of the mechanical part design drawing assignment submitted for the second time is abnormal. At the same time, the instructor's name and contact information are generated in the report, prompting you to find a teacher for guidance.

[0151] In this embodiment, S51 of the method ensures that after the student makes modifications based on the rectification suggestions, the uniqueness of the assignment is verified through the identifier TSid, ensuring that each submitted assignment is associated with the previous version. This mechanism ensures the continuity of the design drawings and the accuracy of the modification records, avoiding the problem of repeated submission or version confusion. S52 re-evaluates the second submitted assignment by re-executing S1 to S3, generates a comprehensive score Tscore2 for the second assignment, and evaluates the improvement effect by calculating the total score difference ΔTscore. When the improvement is effective, the system does not generate a difference report; when the improvement is invalid or an abnormality occurs, the system will automatically generate a design difference report to prompt the student where the problem is, and attach the contact information of the instructor to help the student get further guidance.

[0152] Example 7

[0153] See also Figure 2 , an automatic homework grading system, including a homework submission analysis module, a grading analysis module, a comprehensive grading module, a feedback and optimization suggestion module and a secondary evaluation difference analysis module;

[0154] The job submission analysis module imports the submitted mechanical part design drawing job into the CAD software, extracts the geometric data of the mechanical part design drawing job, and parses the mechanical part design drawing job through the CAD parsing tool to obtain a geometric feature set, and integrates the CAD software with the automatic correction application to extract the acquired geometric data and geometric feature set in real time;

[0155] The correction analysis module performs dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis based on the acquired geometric data and geometric feature set through an automatic correction application, and outputs a dimensional accuracy score Sdim, a geometric structure rationality score Fvalidity, a material selection rationality score Mmatch and a part geometric symmetry score Ssymmctry;

[0156] The comprehensive correction module comprehensively calculates the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, outputs the comprehensive score Tscore of the job, and performs a preliminary comprehensive evaluation on the preset job qualification threshold T and the obtained comprehensive score Tscore of the job, and corrects the qualification of the currently submitted mechanical part design drawing job;

[0157] The feedback and optimization suggestion module generates an identifier TSid after a preliminary comprehensive evaluation. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry.

[0158] After the secondary evaluation difference analysis module modifies the drawing through rectification suggestions, it submits the mechanical part design drawing job for the second time, verifies the uniqueness of the drawing through the identifier TSid, and then executes S1 to S3, outputs the second job comprehensive score Tscore2, calculates the output total score difference △Tscore, and generates a design difference report.

[0159] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for automatically correcting homework, characterized in that: The following steps are involved: S1. Import the submitted mechanical part design drawing job into the CAD software, extract the geometric data of the mechanical part design drawing job, and parse the mechanical part design drawing job through the CAD parsing tool to obtain the geometric feature set, and integrate the CAD software with the automatic correction application to extract the obtained geometric data and geometric feature set in real time; S2. Through the automatic correction application, based on the acquired geometric data and geometric feature set, dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis are performed, and the dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry are output; S3, comprehensively calculate the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, and output the comprehensive score Tscore of the job, and make a preliminary comprehensive evaluation based on the preset job qualification threshold T and the obtained comprehensive score Tscore, and correct the qualification of the currently submitted mechanical part design drawing job; S4. After preliminary comprehensive evaluation, the identifier TSid is generated. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry; S5. After modifying the drawings according to the rectification suggestions, submit the mechanical parts design drawing job for the second time, verify the uniqueness of the drawings through the identifier TSid, and then execute S1 to S3, output the comprehensive score Tscore2 of the second job, calculate the output total score difference △Tscore, and generate a design difference report.

2. The method for automatic homework correction according to claim 1, characterized in that: Said S1 includes S11, S12 and S13; S11, importing the submitted mechanical part design drawing work into the CAD software, then identifying the geometric data in the mechanical part design drawing work through the Measure tool, and exporting the geometric data into a table file; The geometric data include the design size dstudent, the hole diameter Dhole, the shaft diameter Dshaft, the left side length Lleft of the symmetry axis and the right side length Lright of the symmetry axis; S12, automatically analyzing the mechanical parts design drawing operation through the CAD analysis function to obtain the geometric feature set; The geometric feature set includes connection tolerance TOL, geometric structure characteristic coefficient Gshape, and material property characteristic parameter Pmaterial; The connection tolerance TOL is obtained by extracting the hole diameter Dhole and the shaft diameter Dshaft in the table file corresponding to the geometric data and performing tolerance calculation output. The specific algorithm formula is: ; The geometric structure feature coefficient Gshape is directly extracted from the mechanical part design drawing by the CAD parsing tool. The geometric structure features include but are not limited to holes, grooves, bosses and complex surfaces. Then, for each geometric structure feature i, the sub-feature k is obtained according to the functional requirements of the design. The sub-features include but are not limited to the aperture, wall thickness and curvature. The geometric structure feature coefficient Gshape is calculated. The specific algorithm formula is: ;In the formula, Gshape i represents the i-th geometric structure characteristic coefficient, n i represents the total number of geometric structure features of the i-th geometry, f represents the evaluation function, and Xshape i，k Represents the kth sub-feature parameter value of the i-th geometric structure feature, Crequire i，k represents the standard value of the kth sub-feature parameter of the i-th geometric structure feature; The material property characteristic parameter Pmaterial is directly extracted through a CAD analysis tool and matched with a material database, and the material property in the material database is directly used as a preset value; S13. By building an automatic correction application and setting an API application interface to integrate with the CAD software, after the mechanical part design drawing is imported into the CAD software and the geometric data and geometric feature set are obtained, the geometric data and geometric feature set are automatically extracted through the API application interface.

3. The method for automatic homework correction according to claim 2, characterized in that: The S2 includes S21, S22, S23 and S24; S21, through the automatic correction application, based on the acquired geometric data and geometric feature set, extract the connection tolerance TOL and the design dimension dstudent to perform dimensional error analysis, and output the dimensional accuracy score Sdim; The dimensional accuracy score Sdim is calculated by the following algorithm formula: ； In the formula, C represents the total number of design dimensions, represents the standardized error, dstandard represents the standard design size, represents the standardized error of the c-th design dimension.

4. The method for automatic homework correction according to claim 3, characterized in that: S22, by extracting the i-th geometric structure feature coefficient Gshape from the geometric feature set i , through the geometric shape matching algorithm, calculate and output the geometric structure rationality score Fvalidity, and analyze the geometric structure rationality; The geometric structure rationality score Fvalidity is calculated and output by the following geometric shape matching algorithm; ； Where n represents the total number of geometric structure features, wi represents the weight value of the i-th geometric structure feature, and its specific value is set by the user according to the design requirements of mechanical parts.

5. The method for automatic homework correction according to claim 3, characterized in that: S23, by extracting the material property characteristic parameter Pmaterial from the geometric feature set, using the material matching evaluation formula, calculating and outputting the material selection rationality score Mmatch, analyzing the matching degree between the material performance in the mechanical part design drawing operation and the design requirements, and verifying the selection of the material information for the mechanical part design drawing operation; The material selection rationality score Mmatch is calculated and output by the following material matching evaluation formula; ； Where m represents the number of material properties, Pmaterial j Represents the characteristic parameter of the j-th material property, Prequired j represents the material property characteristic parameter required for the j-th design, M j Represents the weight value corresponding to the j-th material property.

6. The method for automatic homework correction according to claim 3, characterized in that: S24, by extracting the left side length Lleft and the right side length Lright of the symmetry axis in the geometric data, calculating and outputting the part geometric symmetry score Ssymmctry, and performing a symmetry rationality analysis on the geometric structure; The part geometric symmetry score Ssymmctry is calculated and output by the following algorithm formula; ； Where Lleft i represents the left symmetric length of the i-th geometric structure feature, Lright i represents the right symmetric length of the i-th geometric structure feature, and n represents the total number of geometric structure features.

7. The method for automatic homework correction according to claim 1, characterized in that: The S3 includes S31 and S32; S31, based on the obtained dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, perform comprehensive calculation and output the comprehensive score Tscore of the job; The comprehensive score Tscore of the assignment is calculated by the following algorithm formula: ； Where a1, a2, a3 and a4 represent the preset weight values ​​of the dimensional accuracy score Sdim, the geometric structure rationality score Fvalidity, the material selection rationality score Mmatch and the part geometric symmetry score Ssymmctry, respectively, and a1+a2+a3+a4=1, and the specific values ​​are set by the user; S32, based on the user-set job qualification threshold T and the obtained job comprehensive score Tscore, a preliminary comprehensive evaluation is performed to analyze the qualification of the mechanical parts design drawing job submitted for the first time. The specific evaluation contents are as follows; When the comprehensive score of the job Tscore ≥ the job qualification threshold T, it means that the mechanical parts design drawing job is qualified; When the comprehensive score Tscore of the job is less than the job qualification threshold T, it means that the mechanical parts design drawing job is unqualified.

8. The method for automatic homework correction according to claim 7, characterized in that: The S4 includes S41 and S42; S41, after preliminary comprehensive evaluation, assign a unique identifier TSid to each mechanical parts design drawing job, and after setting the identifier TSid, transfer and save it to the automatic correction application database; The identifier TSid is generated in the following way: , where TimeStamp represents the timestamp, CourseID represents the course ID, and StudentID represents the student ID; S42. When the preliminary comprehensive assessment shows that the mechanical part design drawing is unqualified, the initial set size requirement index Cc, geometric structure rationality requirement index Hl, material requirement index Cl and symmetry requirement index Dc are automatically compared with the size accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, respectively, to obtain the output results and generate modification suggestions. The specific modification suggestions are as follows; When the dimensional accuracy score Sdim≤dimensional requirement index Cc, it meets the requirements and no prompt is needed; When the dimensional accuracy score Sdim>dimensional requirement index Cc, it does not meet the requirements and is marked as needing adjustment. The modification suggestion prompts to optimize the dimensional accuracy. When the geometric structure rationality score Fvalidity ≥ geometric structure rationality requirement index Hl, the requirement is met and no prompt is needed; When the geometric structure rationality score Fvalidity is less than the geometric structure rationality requirement index Hl, it does not meet the requirements and is marked as needing adjustment. The rationality of optimizing the geometric structure is prompted in the modification suggestions; When the material selection rationality score Mmatch≥material requirement index Cl, it meets the requirements and no prompt is needed; When the material selection rationality score Mmatch is less than the material requirement index Cl, it does not meet the requirements and is marked as requiring adjustment. In the modification suggestions, it is suggested to optimize the rationality of material selection; When the part geometric symmetry score Ssymmctry = symmetry requirement index Dc, the part symmetry meets the requirements and no prompt is needed; When the part geometric symmetry score Ssymmctry is less than the symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometric structure symmetry; When the part geometry symmetry score Ssymmctry>symmetry requirement index Dc, the part symmetry does not meet the requirements, and the modification suggestion prompts to optimize the geometry symmetry.

9. The method for automatic homework correction according to claim 8, characterized in that: The S5 includes S51 and S52; S51. After modifying the mechanical part design drawing job according to the rectification suggestion, submit the mechanical part design drawing job for the second time, and verify the uniqueness of the mechanical part design drawing job by inputting the identifier TSid into the automatic correction application; S52, after verifying the uniqueness, execute S1 to S3 for the second time, output the second job comprehensive score Tscore2, and perform difference calculation with the job comprehensive score Tscore to output the total score difference △Tscore, and generate a design difference report based on the total score difference △Tscore output result; The total score difference △Tscore is calculated and output by the following algorithm formula; ； The design difference report is specifically generated by the following contents: When the total score difference △Tscore>0, it means that there is improvement in the second submission of mechanical parts design drawing work, and no difference report is generated at this time; When the total score difference △Tscore≤0, a difference report is generated, indicating that the modification of the mechanical part design drawing assignment submitted for the second time is abnormal. At the same time, the instructor's name and contact information are generated in the report, prompting you to find a teacher for guidance.

10. An automatic homework grading system, applied to an automatic homework grading method according to any one of claims 1 to 9, characterized in that: It includes homework submission analysis module, correction analysis module, comprehensive correction module, feedback and optimization suggestion module and secondary evaluation difference analysis module; The job submission analysis module imports the submitted mechanical part design drawing job into the CAD software, extracts the geometric data of the mechanical part design drawing job, and parses the mechanical part design drawing job through the CAD parsing tool to obtain a geometric feature set, and integrates the CAD software with the automatic correction application to extract the acquired geometric data and geometric feature set in real time; The correction analysis module performs dimensional error analysis, geometric structure rationality analysis, annotation material information selection verification and geometric structure symmetry rationality analysis based on the acquired geometric data and geometric feature set through an automatic correction application, and outputs a dimensional accuracy score Sdim, a geometric structure rationality score Fvalidity, a material selection rationality score Mmatch and a part geometric symmetry score Ssymmctry; The comprehensive correction module comprehensively calculates the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry, outputs the comprehensive score Tscore of the job, and performs a preliminary comprehensive evaluation on the preset job qualification threshold T and the obtained comprehensive score Tscore of the job, and corrects the qualification of the currently submitted mechanical part design drawing job; The feedback and optimization suggestion module generates an identifier TSid after a preliminary comprehensive evaluation. When the preliminary comprehensive evaluation shows that the mechanical part design drawing is unqualified, the feedback mechanism is automatically triggered to provide specific rectification suggestions by separately analyzing the output dimensional accuracy score Sdim, geometric structure rationality score Fvalidity, material selection rationality score Mmatch and part geometric symmetry score Ssymmctry. After the secondary evaluation difference analysis module modifies the drawing through rectification suggestions, it submits the mechanical part design drawing job for the second time, verifies the uniqueness of the drawing through the identifier TSid, and then executes S1 to S3, outputs the second job comprehensive score Tscore2, calculates the output total score difference △Tscore, and generates a design difference report.

Images