Structured processing and application system for automobile standard laws and regulations

Abstract

The invention relates to the technical field of data processing, in particular to an automobile standard regulation structured processing and application system. The system comprises a product project conformity management and control unit which determines a matched first standard regulation set according to attribute information; extracting keywords in the test process of the test report, and determining a second standard regulation set in the first standard regulation set through a keyword matching algorithm; determining a third standard regulation set in the second standard regulation set according to the step sequence of the matched keyword and the matching of the step content; determining a fourth standard regulation set in the third standard regulation set by performing similarity comparison on the test diagrams; and according to the test requirements of each standard regulation in the fourth standard regulation set, determining whether the target vehicle type meets the standard or not. According to the embodiment of the invention, automatic processing of the standard regulation and automatic and accurate comparison with the test report are realized.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a structured processing and application system for automobile standards and regulations. Background Art

[0002] At present, the automotive industry's management of domestic and international standards and regulations mainly relies on traditional manual processing methods, including manual retrieval, interpretation and application of regulatory documents, resulting in the following problems in existing technologies: 1. Inefficiency and prone to human errors, especially when dealing with a large number of regulatory documents and frequently updated regulatory requirements; 2. Lack of effective structured processing methods, resulting in regulatory clauses being difficult to quickly locate and compare, and lack of effective comparative analysis methods for differences between different versions of regulations, which increases the difficulty of understanding and application; 3. Information islands: Poor information flow between different departments leads to delays in regulatory updates and applications; 4. Compliance risks: Lack of effective tools for regulatory compliance checks increases the possibility of companies facing legal risks.

[0003] In view of this, this application is filed. Summary of the invention

[0004] To solve the problems in the background technology, the present application provides a structured processing and application system for automobile standards and regulations, including: a basic data control unit for managing and maintaining standards and regulations in the automobile field; A standard structured processing unit for structured processing and storage of the standard regulations; An input unit receives attribute information and a test report of a target vehicle model; The product project compliance control unit determines a matching first set of standards and regulations according to the attribute information; extracts keywords from the test process of the test report, and determines a second set of standards and regulations in the first set of standards and regulations through a keyword matching algorithm; determines a third set of standards and regulations in the second set of standards and regulations according to the matching of the step sequence and step content of the matching keywords; determines a fourth set of standards and regulations in the third set of standards and regulations by performing a similarity comparison on the test images; and determines whether the target vehicle model meets the standards according to the test requirements of each standard regulation in the fourth set of standards and regulations.

[0005] This application solves the problems of low efficiency in standard text processing, analysis, and interpretation, information islands, increased compliance risks, and high responsibility in the enterprise standard management process in the prior art. Compared with the prior art, the beneficial effects of this application are: 1. Improved efficiency and accuracy of regulatory processing: The system realizes the rapid online compilation and structured processing of automotive standards and regulations, reduces manual operations, avoids human errors, and significantly improves work efficiency. 2. Enhanced regulatory difference comparison and analysis capabilities: Through the standard difference comparison method, the system can effectively identify and compare the differences between different versions of regulations, simplifying the process of understanding and applying regulations. 3. Promoted information sharing and collaboration: Eliminated the information island phenomenon, ensured the synchronization of regulatory updates and applications between different departments, and enhanced the efficiency of cross-departmental communication and collaboration. 4. Reduced compliance risks: Provided a regulatory compliance investigation function, which helps enterprises to promptly discover and correct possible legal risks and ensure product compliance. 5. Optimized the full life cycle management of enterprise standards: The system covers the entire process of enterprise standards from formulation to revision to abolition, improving the systematicness and efficiency of standard management, and ensuring the continuity and consistency of standards. 6. Strengthened the compatibility of internal and external standards and regulations: ensured that corporate standards can meet industry norms and adapt to regulatory requirements of specific markets, and improved the scope of application and flexibility of corporate standards. 7. Promoted the digital transformation of the automotive industry: through the application of automated and intelligent processes, it promoted the efficient interpretation, analysis and response of automotive companies to standard and regulatory data, accelerated the process of digital transformation, and improved overall operational efficiency and market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0007] Figure 1 It is a structural schematic diagram of a system for structured processing and application of automobile standards and regulations provided in an embodiment of the present application; Figure 2 is a schematic diagram of the structure of a standard structured processing unit provided in an embodiment of the present application; Figure 3 is a schematic diagram of the structure of an OCR text recognition tool based on automotive standards provided in an embodiment of the present application; Figure 4 It is a schematic diagram of the structure of a standard text structured splitting tool provided in an embodiment of the present application; Figure 5It is a structural diagram of a product project compliance control unit provided in an embodiment of the present application; Figure 6 It is a structural diagram of a standard full life cycle management unit provided in an embodiment of the present application. DETAILED DESCRIPTION

[0008] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0009] The present application embodiment provides a structured processing and application system for automobile standards and regulations, which is applicable to the management of standards and regulations in the domestic and foreign automobile industries. Figure 1 The system provided in this embodiment includes: a basic data control unit, a standard structured processing unit, an input unit and a product project compliance control unit. Each unit is introduced below.

[0010] The basic data control unit manages and maintains the standards and regulations in the automotive field. Specifically, the basic data control unit is used to store, access and manage domestic and foreign standards, regulations and policy data, and enterprise standard data, including domestic standards and regulations library, overseas standards and regulations library, domestic and foreign policy library, enterprise standard library and dynamic information library. For example, the domestic / overseas standards and regulations library mainly manages all domestic and foreign standards, regulations and policies collected by the enterprise, including domestic certification standards, domestic mandatory standards, overseas certification standards, etc. The enterprise standard library mainly manages all standard resources formulated within the enterprise, including technical basic standards, product standards, engineering verification technical standards, process technical standards, quality control technical standards, etc. The above database mainly realizes the entry storage and detail display functions, and can customize the configuration system directory, data management and maintenance, data import and export, text online preview, text multi-language translation, intelligent query, automatic status change, data issuance and borrowing, providing value-added information, and authority control. In particular, in addition to the above functions, the enterprise also associates the enterprise standard plan, revision, review, and abolition processes to achieve dynamic update of the standard status. The various functions of the above database are described in detail below.

[0011] 1. The entry-based storage and detail display function is used to store and display standard regulations, policies, and corporate standards in an entry-based manner, and you can view the detailed information of each standard regulation. The attribute fields of standard regulations and policies include basic standard information (standard number, standard name, implementation date, applicable certification, energy type, applicable vehicle model, applicable region, and other information inherent in the standard itself), extended information (responsible department, responsible engineer, related assemblies and parts, related test items, and other corresponding information within the enterprise, as well as interpretation documents related to standards and regulations), text information (release drafts, process drafts, supplements, etc.), and attachment value information (standard interpretation, standard decomposition sheet, comparison results, question and answer records, project compliance assessment results, etc.). The attribute fields of enterprise standards include basic information of the enterprise standard (enterprise standard number, enterprise standard name, confidentiality level, standard status, release date, implementation date and other information inherent in the standard itself), drafting information (drafting person, drafting department, applicable vehicle model, energy type, system category, related assembly and component numbers, etc.), text information (release version, historical version, preparation instructions, amendment sheet, etc.), and related information (replacement enterprise standard number, replaced enterprise standard number, referenced standards, related processes, related files, etc.).

[0012] 2. The custom system classification function is used to customize the standard and regulatory system directory and display the standard, regulatory and policy data under different categories in a tree structure.

[0013] 3. Data management and maintenance functions are used to add, edit, delete, configure, remove, favorite, and share standard and regulatory data (including detailed information and related attachments). The add operation is used to add a new standard regulation, the edit operation is used to modify the detailed information and related attachments of a standard regulation that already exists in the database, the delete operation is used to delete the standard regulations in the database individually or in batches, the configure operation is used to add standard and regulatory data in batches to the selected directory, the remove operation is used to remove standard and regulatory data in batches from the selected directory, the favorite operation is used to add a standard regulation to the personal center unit - My Favorites, and the share operation is used to share a standard regulation with other users and view it in the collected shares of the personal center unit.

[0014] 4. Data import and export function is used to provide standard basic data EXCEL import template. After entering the standard and regulatory details according to the filling rules, the local standard and regulatory policies or enterprise standard data can be imported into the corresponding directory in batches by uploading the template and related attachment texts. At the same time, the data already in the database can be selected and exported in batches, including detailed information and corresponding attachments.

[0015] 5. The online file preview function is used to support the online preview of the corresponding attachments (pdf, docx, excel format) in the detailed information of standards and regulations. Personal notes can be added in the preview, and a watermark can be added to the global preview page to increase confidentiality.

[0016] 6. Text multi-language translation function, which is used to convert foreign standards into Chinese text through big data translation engine and provide preview and download.

[0017] 7. Intelligent query function, supports general query by attribute fields such as standard number, standard name, and multi-condition combination query.

[0018] 8. The standard status automatic change function is used to determine the status of each standard and regulation according to the standard status change rules, the current time and the substitution relationship between standards, including: drafting, soliciting opinions, review, approval, release, upcoming implementation, currently valid and abolished. Specifically, it is automatically updated through fields such as replacement standards, implementation date and process progress. The specific standard status change rules are: ① When the standard is put into the library, you can manually select the release and its previous status; ② When the standard status is "released", the next day, the standard that has been released will be judged to see whether the current date has reached the implementation date. If it has not reached the implementation date, it will be displayed as "to be implemented", and if it has reached the implementation date, it will be displayed as "currently valid"; ③ For the currently valid standards, they can be replaced by new standards and automatically changed to abolished (when the new standard status is not "currently valid", the old standard status is still "currently valid", and when the new standard status is "currently valid", the replaced standard status will become abolished. ), or it can become abolished after reaching the abolition date. After the standard is abolished, it is not allowed to change its status; the status of enterprise standards are: planned revision, released, under revision, revised, and abolished. The specific standard status change rules are: enterprise standard A replaces enterprise standard B. When the application process for the formulation and revision of enterprise standard A is approved, enterprise standard B becomes planned revision; enterprise standard A replaces enterprise standard B; when the formulation and revision process of enterprise standard A enters the approval stage, the standard status of enterprise standard B automatically becomes "under revision". When the formulation and revision process of enterprise standard A is completed, enterprise standard A is published and enterprise standard B becomes revised.

[0019] 9. The data issuance and borrowing function is mainly used to grant users who do not have the right to view standard details and preview attachments the permission to view detailed information preview and download related attachments. Users can apply for borrowing permissions for multiple standards under a certain category.

[0020] 10. Provides value-added information functions, supports the update log (log file), interpretation report, split results, comparison results, standard questions and answers, compliance judgment results and other modules of the data in the standard and regulation details, and can jump to the corresponding module with one click to view the application status of the corresponding standard. In particular, for enterprise standards, you can jump to the corresponding enterprise standard formulation and revision process and view the enterprise standard score.

[0021] 11. The permission control function is used to control the user's data access rights and menu function permissions based on organizational levels, roles and other dimensions, and is divided into data permissions and functional permissions. Among them, data permissions include directory-level permissions and single standard permissions, which are used to set a certain type of user to only access a certain directory or data with a certain characteristic attribute. In particular, for the enterprise standard library, different detail viewing permissions can be set according to the enterprise standard confidentiality level: everyone with an enterprise standard confidentiality level of C1 can view the details, only the drafting department and related department personnel can view C2, and only the main drafter, other drafters, enterprise standard engineers, system administrators and those who apply for authorization separately can view C3. Functional permissions include module permissions and button permissions.

[0022] See also Figure 2 , standard structured processing unit, performs structured processing and storage of standards and regulations. This unit includes: OCR text recognition tools based on automotive standards, standard text structured splitting tools, standard text difference comparison tools and standard online editing tools, forming four aspects of processing and application: external standard data format conversion, external standard itemization decomposition, interpretation of differences between old and new external standards, and online formulation and revision of enterprise standards. The structured data generated by this unit constitutes a structured database of standards and regulations, which serves as information input for the product project compliance management unit. The standard online editing tool can be applied to the enterprise standard full life cycle management unit to carry out online formulation and revision of enterprise standards. The functions implemented by each tool are introduced below.

[0023] 1. The OCR (Optical Character Recognition) text recognition tool based on automotive standards is mainly used to convert the non-editable domestic and international standard and regulatory resources uploaded by users into editable files. After the user enters the tool, click Upload to upload the PDF file to the server. The service uploads the user's uploaded file to the OCR recognition cloud service through the webserver. In the cloud service, the system will automatically recognize and convert the PDF file, and finally convert the uploaded PDF file into an editable Docx document. See Figure 3 The tool consists of a central processing module, an image preprocessing module, an OCR recognition module, a file conversion module, a user account management module, and an external interface. The specific working steps include: S101. File upload and reception: Users log in to the standard and regulation structured application system through the user account management module, enter the tool to upload files, call the server external interface to upload PDF files, the external interface receives files and saves them to the file directory, queues the task request for processing, and saves the user and file information to the database.

[0024] S102, file preprocessing: the central processing module takes out the tasks to be processed from the task queue, and sends the files to the image preprocessing module for processing, extracts images from the PDF file and uses OpenCV to adjust the image size, reduce noise, enhance contrast, adjust grayscale, binarize, etc.

[0025] S103, OCR text recognition and format conversion: The pre-processed files are sent from the central processing module to the OCR recognition module. Based on the big data deep learning algorithm based on the Abbyy12 algorithm core library, the OCR SDK is used to perform text image recognition on the files, and then the recognized content is passed through the file conversion module to generate a Docx file.

[0026] S104, post-processing correction: the file conversion module continues to perform intervention correction on some Chinese characters with recognition errors that appear regularly in the converted file.

[0027] S105, file output: The corrected file is stored in the database by the central processing module, and a callback is performed to return the final file to the standard and regulation structured application system through the external interface.

[0028] 2. Standard text structured splitting tool, which splits the text, tables, pictures and other information in the editable text content according to the clause structure to form structured data, and reorganizes, arranges and displays the structured data according to the clause tree structure, forming editable knowledge items in terms of clauses, which are stored in the standard and legal structured database. The standard and legal structured database can perform attribute maintenance and share collections of knowledge items, and is associated with the intelligent retrieval center. In the process of searching for useful knowledge, it is only necessary to retrieve and analyze the knowledge items to filter out more fine-grained retrieval results. The tool consists of a data reading module, a data rule analysis module, a structured data generation module, and a data storage and display module. Its structure diagram is shown below. Figure 4 As shown. Combined with the actual application scenarios and effects, the specific working steps of the standard text structure splitting tool include: S201. File upload and reception: The user logs in to the standard and regulation structured application system, enters the standard text structured splitting tool, uploads the standard and regulation file in doc / docx format stored locally or generated by the OCR text recognition tool, and reads the standard number / policy number, standard name / policy name and text content in the file through the data reading module.

[0029] S202. For each standard regulation, confirm the splitting range, for example, filter out all target paragraphs within the range according to the chapter number.

[0030] S203, traverse each paragraph in the split range, and mark the paragraphs that meet the clause identification rules with clause numbers. Read the clause identification library, traverse each paragraph at the same time, mark the corresponding clause numbers for the paragraphs that meet the clause identification rules, obtain the detailed content and enter step S204, then loop through each paragraph in the split range until each paragraph is marked with a clause number.

[0031] Among them, the paragraphs of the symbolic clause identification rules generally contain any one or a combination of two or more of the three elements of multi-level document numbering, chapter title and chapter content.

[0032] S204, identify paragraphs with the same clause number and merge them, and sort the merged paragraphs according to the context order and clause level, and extract and split the content in terms of clauses. Specifically, identify paragraphs with the same clause number and merge them, and then sort the merged paragraphs according to the context order and clause level, and extract and split the content in terms of clauses. During the extraction, the data is divided into three types according to the data type definition: text type, table type, and image type, and the data is split and converted according to the different data types. For text data, paragraph separation storage and key attribute analysis are performed, and attribute words are marked. For table data, two-dimensional structure storage is performed and the text position is marked. For image data, it is stored in the form of attachments, and then different types of data are arranged and assembled according to the order in the clauses, and text, table and image data are respectively used as the minimum storage unit for structured storage.

[0033] S205, Data storage, synchronization and maintenance: Display the structured text in the form of a tree directory, with the directory levels being chapter, first-level clause number, second-level clause number, ... N-level clause number. After the split is completed, it is automatically synchronized to the detailed information of the corresponding standard of the basic data control unit. The clauses can be maintained with attribute tags to associate vehicle assemblies and parts, responsible departments, fill in the applicable target market, vehicle type, power type, etc., and the text, tables, and pictures in the clauses can be edited again and referenced to other standards.

[0034] 3. The standard text difference comparison tool, based on the structured data generated by the standard text structured splitting tool, performs difference analysis and marking on different standards and regulations, and generates a standard and regulation difference interpretation report, so as to quickly identify the difference information between different versions of standards and regulations. In an application scenario, the difference comparison between two similar or previous and subsequent versions of standards and regulations is carried out, and the specific working steps include: S301 , preprocessing two standards and regulations, for example, preprocessing the texts, tables, and pictures of the two input standards and regulations.

[0035] Preprocessing of text data: word segmentation and standardization, stop word filtering and stem extraction, and construction of the semantic tree structure of regulatory clauses. Word segmentation and standardization is to segment the regulatory text and standardize the format of regulatory terms to ensure the consistency of subsequent analysis, especially the special terms in the automotive industry (such as "ABS", "ESP", etc.); stop word filtering and stem extraction remove irrelevant words, extract stems, and reduce the impact of synonyms; semantic tree construction is to build a semantic structure tree of regulatory clauses based on grammatical analysis to facilitate deeper semantic understanding.

[0036] Preprocessing of tabular data: convert the table content into a structured data format, such as XML or JSON format, to facilitate further data processing; standardize the numerical data in the table to ensure the uniformity of the data format, such as standardization of units (such as kilometers / hour, millimeters, etc.); process the text content in the table in the same way as plain text data.

[0037] Preprocessing of image data: For the text content in the image, use the OCR text recognition tool to identify it and convert it into an editable text format; for the charts or complex symbols in the image, including structural diagrams, circuit diagrams, etc., perform ① binarization processing: convert the image to black and white to simplify the subsequent processing steps; ② edge detection: use the Canny edge detection algorithm to detect the lines and boundaries in the image; ③ shape recognition: identify specific shapes (such as circles, rectangles, polygons, etc.) to determine the position and type of chart elements; ④ pattern recognition: use deep learning technology to identify specific chart types or complex symbols; ④ feature extraction: identify and extract key features such as part numbers and connecting lines in the chart; ⑤ data conversion converts the extracted information into a graphic model SVG file to maintain the vector characteristics of the graphic.

[0038] S302. Use a deep learning model to perform semantic analysis on the preprocessed standard regulations to obtain two matching clauses. Specifically, use a deep learning model, such as BERT or Transformer, to perform semantic analysis on the preprocessed text, convert the text into a semantic vector, and calculate the cosine similarity between the vectors to achieve accurate matching of different clauses across standard regulations.

[0039] S303, use the distance algorithm to perform a preliminary difference analysis on any two clauses. For example, the Levenshtein distance is an indicator that measures the difference between two sequences by calculating the minimum number of single-character editing operations (including insertion, deletion or replacement) required to change from one sequence to another. For text data: use the Levenshtein distance algorithm to calculate the edit distance between texts to quantify the similarity between the two texts, and record the value as a preliminary quantitative difference indicator; for table data: calculate the Levenshtein distance of each cell, and comprehensively consider the structural differences of the entire table. If the table structure has changed, these differences can be identified by comparing the number and order of rows and columns; for image data: use the editable file preprocessed in S301, and also use the Levenshtein distance algorithm to calculate the difference in the converted text. If there is content that is difficult to convert, the system will store the compared pictures in the manual verification library for manual difference comparison. After preliminary difference analysis, two relatively similar clauses can be obtained.

[0040] S304, using the Smith-Waterman (SW) algorithm and the diff algorithm, perform an in-depth difference analysis on the two clauses. Specifically, for the two relatively similar clauses obtained from the preliminary difference analysis, the SW algorithm is first used to perform a local sequence alignment on the two clauses to identify the maximum matching fragment in the text, and to help identify similar but not identical clause paragraphs in the standard regulatory text. This is applicable to clauses that have been slightly modified but whose core meanings have not changed and for which the differences cannot be identified in S303. Then the diff algorithm is used to mark the text differences, clearly showing which content has been added, deleted or modified, and highlighting the specific changes in the standard regulatory texts. This embodiment combines the SW algorithm with the Levenshtein algorithm to identify the local maximum matching areas and calculate the differences between these areas, while the diff algorithm is used to assist in displaying the specific locations and properties of these differences. The three complement each other and jointly complete a detailed difference analysis.

[0041] S305, quantify and mark the preliminary difference analysis and the deep difference analysis, and generate a difference report. Specifically, the specific location and nature of the change are accurately marked by quantitative methods, and combined with the results of the above steps, each identified difference is quantitatively evaluated, such as calculating the similarity percentage, and the location of the difference is marked in the text with specific marks (such as color coding, comment box) to facilitate intuitive understanding by users.

[0042] 4. Standard online editing tool, which enables users to edit standard documents online, preset the document format template required by the enterprise, produce documents in a fixed format, and quickly generate enterprise standard texts that meet the writing specifications. Through the custom editor, it is possible to insert tables, pictures, directories, appendices and other contents in the document preparation process, and support the acquisition of standard information content through other module interfaces in the system and directly insert it into the enterprise standard file, and finally form an enterprise standard library. The core functions of this tool include: 1) Initialize the enterprise standard template: When entering the enterprise standard online editing, you can select the corresponding enterprise standard prefabricated template according to the enterprise standard type, initialize the document content after selection, automatically generate the preface, directory and other information, and preset the format of some chapters, clauses, texts and other formats according to the requirements of the enterprise. The user input content defaults to the overall document requirements, and the format can be modified according to the actual business scenario. 2) Automatic filling of content: When compiling the content of the enterprise standard, some content such as standard name, standard number, replacement situation, implementation date, compilation department, etc. can be automatically brought in from the preparation and revision process data and filled in to the required position. 3) Automatic numbering of chapters and clauses: When users add new chapters, sections and clauses, they insert content for automatic numbering. 4) Local file export: You can export the same Word version file as the one edited online. 5) Provide format brush: Support copying text. By clicking the shortcut keys of chapter, article, note, paragraph, etc., the copied text can be brushed into the corresponding text content format, which is convenient for user operation. 6) Support structured splitting: After the user finishes writing, it can be passed to the standard text structured splitting tool for structured disassembly. After the disassembly is completed, it is bound to the enterprise standard and passed to other business systems.

[0043] The system also includes an input unit for receiving attribute information and a test report of a target vehicle model. The attribute information includes vehicle model information, test item information, region information (the region where the target vehicle model is sold) and time information (the time when the target vehicle model is launched). The test report is a report generated by testing the target vehicle model, including a test process, a test diagram and a test report result.

[0044] The product project compliance control unit determines the first set of matching standards and regulations based on the attribute information. Specifically, the first set of standards and regulations that are valid within the geographical information and time information and that meet the vehicle model information and test item information are determined based on the attribute information. The first set of standards and regulations includes multiple standards and regulations. Subsequently, the product project compliance control unit uses a series of algorithms to find the standard regulations that match the test items in the first set of standards and regulations, specifically including the following steps: Step 1: extract keywords from the test process of the test report, and determine the second standard regulation set from the first standard regulation set through a keyword matching algorithm.

[0045] Keywords can be the name of the test item, the name of the test equipment, test parameters and indicators, etc. Keywords are extracted from the first standard regulation set and the test process, and matched in pairs, so as to obtain a second standard regulation set with keyword matching. The second standard regulation set is selected from the first standard regulation set.

[0046] Step 2: Determine a third standard regulation set in the second standard regulation set according to the matching of the step sequence and the step content where the matched keywords are located.

[0047] The vehicle testing process includes multiple steps. The order of some steps can be swapped. For example, there is no substantial difference between adjusting the temperature of the environmental chamber first and then adjusting the humidity of the environmental chamber and adjusting the pressure of the environmental chamber first. However, there is a big difference in whether the step content is high temperature or low temperature. Based on this, this embodiment puts the matching keywords back to the original steps, so that matching steps are obtained. In these matching steps, the influence of the step order is weakened and the influence of the step content is strengthened. Then, the following formula is used to calculate the matching M of the step order and step content of the matching keywords in the second standard regulation set and the test report; the standard regulations with M greater than the set value are added to the third standard regulation set. The set value can be set by yourself.

[0048] ; Among them, A and B are coefficients, and , i is the index of the keyword, N is the number of keywords, It is a key word in the standard regulations The steps are It is a key word in the testing process The steps are and is the matching keyword, express and The difference in the order of steps, for example, It is step 5. is step 6, then =6-5=1. express and The semantic distance of the step contents can be expressed, for example, by semantic similarity.

[0049] The third standard regulation set is selected from the second standard regulation set.

[0050] Step 3: Determine the fourth standard regulation set in the third standard regulation set by performing similarity comparison on the test images.

[0051] Considering that some test contents are not written in the standard regulations and test reports in text form, but reflected in the test drawings. For example, a chassis dynamometer is drawn in the test drawing, but since the chassis dynamometer is a conventional device, it is not reflected in the steps of the test report. Therefore, it is necessary to compare the similarity of the test drawings in the standard regulations and the test report. The test drawing can be a mechanical drawing or a site photo. In this step, if the above keywords and steps are matched, the test drawings are further compared to see if they are similar.

[0052] Optionally, search for test images around the steps where the matching keywords are located; use an image similarity comparison algorithm and a keyword comparison algorithm for the test image annotation names to determine the fourth standard regulation set in the third standard regulation set. For example, the test image is converted into a short, fixed-length fingerprint through a hash algorithm, and the similarity comparison of the test image is achieved through fingerprint comparison. Perform keyword comparison on the test image annotation names (such as the name of each labeled test component). If the test image and annotation name of a standard regulation in the third standard regulation set match the test image and annotation name in the test report, then the standard regulation is added to the fourth standard regulation set.

[0053] Step 4: Determine whether the target vehicle model meets the standards according to the test requirements of each standard regulation in the fourth standard regulation set.

[0054] Basically, if the standards and regulations in the fourth standard and regulation set match the test report, the project compliance judgment process will automatically assign responsible persons for investigation of each standard and regulation in the fourth standard and regulation set that the target vehicle model needs to meet, create assessment tasks, and execute assessment tasks; and determine whether the test report results meet the qualitative and quantitative requirements of each standard and regulation in the fourth standard and regulation set.

[0055] The function of the product project compliance control unit is explained below through a more detailed embodiment.

[0056] like Figure 5As shown in the figure, the product project compliance control unit includes a vehicle project library, a vehicle development compliance list, a project compliance determination process (project line / regulatory line), an analysis of standard-related projects, a non-conforming item tracking library, and a non-conforming item rectification process. This unit uses domestic and foreign standard and regulatory libraries and a structured database of standards and regulations as inputs. First, the regulatory requirements are intelligently matched with the target vehicle information to form a development compliance list for the target vehicle. Then, through the project compliance determination process, the responsible person for investigation is automatically assigned to each regulatory requirement that the target vehicle needs to meet, an assessment task is created, the assessment task is executed, and the automotive product design parameters (such as vehicle basic attributes, performance parameters, component installation parameters, etc.) are automatically or manually determined to meet the quantitative and qualitative requirements in the regulatory provisions. All assessment results (compliant, non-compliant, not involved, to be verified) generated during the assessment process are synchronized to the standard-related project analysis and vehicle development compliance list. Non-compliant problem items (non-compliant, to be verified) in the assessment results will be synchronized to the non-conforming item tracking library for tracking, and finally the problem rectification and closure will be completed through the non-conforming item rectification process. The execution process of the product project compliance control unit includes the following operations: S401, obtain, create and edit vehicle model project information. Engineers automatically obtain or upload project documents in the vehicle model project library to automatically parse vehicle model information, quickly create vehicle model projects, generate original vehicle model data, and customize and edit the original data according to business needs to input vehicle model attribute information for subsequent compliance determination processes.

[0057] The vehicle model project library can synchronize the vehicle information of the domestic market, overseas export, under-development and in-production within the enterprise according to the needs, including the main page, detail information page and vehicle model development compliance list. The main page stores and manages vehicle model data in the form of a list, and can automatically create, edit, retrieve and export lists of vehicle model projects; the vehicle model project details page is used to display all the basic data and technical parameters of the vehicle model after creation. The system can set viewing permissions for different roles, and users with permissions can view detailed information; the vehicle model development compliance list is used to generate a list of standards and regulations that the vehicle model needs to meet, display the judgment results synchronized from the project compliance judgment process, and summarize the number of risk items of different levels (red, yellow, green).

[0058] When engineers create vehicle model projects, in order to reduce the workload of manually screening and filling in attribute information when creating vehicle model projects, improve data processing efficiency, and reduce the risk of manual errors, the vehicle model project library provides vehicle model data automatic synchronization and parsing functions, mainly including synchronization and parsing rule settings and automatic synchronization mechanism settings. Specifically, 1) Synchronization and parsing rule settings: According to the needs of compliance investigation, configure synchronization rules (including vehicle model requirements, attribute field ranges), and extract only a certain type of vehicle model (for example, selectively synchronize a certain type of vehicle model according to dimensions such as vehicle model category, power type, and configuration level) and key vehicle model attribute information used for compliance judgment (for example, only synchronize vehicle type, active safety, and passive safety technical parameters for compliance judgment in the safety field); 2) Automatic synchronization mechanism settings: Set the synchronization frequency in advance, such as once a day or a week, or when the project reaches a certain project node, you can achieve scheduled automatic synchronization to ensure that the vehicle model attributes used in compliance investigation are the latest. The vehicle attribute information created after extraction generally includes basic data (such as the target market (region) of the vehicle model, project code, project name, project members, vehicle category (M1, M2, M3, N1, N2, etc.), power type (gasoline, diesel, HEV, PHEV, etc.), announced vehicle model, valve time nodes of each project, etc.) and technical parameters (appearance design, structural design, interior design, performance design, testing and verification, etc.).

[0059] S402: Assign project compliance assessment roles. Engineers configure the permissions of various roles involved in the project compliance assessment process within the project, and users with corresponding role permissions perform different node tasks in the subsequent compliance assessment process.

[0060] The authority configuration of each role can be customized according to the company's own situation; usually, the regulatory engineer or project leader is the initiator of the process and can issue assessment tasks to various professional fields; the interface person or person in charge of each professional field is the distributor of compliance determination tasks, and according to the situation of their own professional field, the regulations or clauses that need to be evaluated are distributed to the subordinate design engineers; the design engineer receives the task and fills in the compliance determination results of the corresponding regulations or clauses based on the current situation of the project; the experts or managers of each profession are the reviewers of the content of the regulatory compliance determination results.

[0061] S403, automatically generate the initial vehicle model development compliance list. After completing the creation of the vehicle model project and the assignment of project roles, the system automatically associates and matches the external standard regulations and their structured clauses with the vehicle model project based on the target market, vehicle type, power type and other attribute conditions, and specifically forms a preliminary matching table between the vehicle model product and regulatory requirements by combining the fourth standard regulation set with the test report.

[0062] S404. Execute the vehicle model project compliance determination process (including qualitative and quantitative requirements) according to the preliminary matching table, automatically assign and issue the determination responsible person, and fill in and review the compliance determination results. The regulatory engineer selects the compliance determination process type, further screens the vehicle model development compliance list to form a compliance determination task list, and automatically assigns the field interface person / system engineer according to the professional fields involved in the regulatory requirements. The field interface person can continue to distribute tasks, and the design engineer determines the compliance of the received regulatory requirements according to the actual project status. The quantitative requirements can be automatically determined. After the determination results are submitted, they are reviewed by experts from various departments / professions to complete the compliance assessment of the project.

[0063] In a specific implementation, the regulatory engineer initiates the project compliance determination process in the task processing unit. After the process is initiated, he needs to select any one of the following two lines: one is the project line, which reviews and reviews the regulations in real time at each node from before to after the launch, and makes compliance determinations on all regulatory requirements that need to be met by this vehicle model project; the other line is the regulatory line, which identifies all vehicle model projects that may be involved for compliance determination when new regulations are released or updated. Regardless of whether it is the project line or the regulatory line, after the compliance determination process, the latest regulatory requirements and corresponding evaluation results will be updated to the vehicle model development compliance list to ensure the timeliness and completeness of the list. The vehicle model project compliance determination process includes the following operations: S4041. Select the process type to generate different types of compliance determination task lists, including project line compliance determination task lists and regulatory line compliance determination task lists. Specifically, if the project line compliance determination process is selected, select a vehicle model in the vehicle model project library and select the fourth standard and regulation set to generate a single vehicle model-multiple standard compliance determination task list. If the regulatory line compliance determination process is selected, select a newly released or updated regulation in the domestic and foreign standard and regulation libraries, the system retrieves the R&D compliance list data of each vehicle model, automatically selects the vehicle models that have included the regulation, and generates a single standard-multiple vehicle model compliance determination task list.

[0064] S4042. Automatically assign professional interface persons to the standards and regulations in the compliance determination task list and issue determination tasks. Specifically, after the compliance determination task list is formed, the system automatically brings out the corresponding professional field person in charge based on the associated assemblies, parts and responsible departments maintained on the standards / clauses. For example, if the associated assemblies and parts of a certain standard in the list are engine assemblies and cylinder blocks, and the responsible department is the power system R&D department, the system automatically reads the personnel position information from the enterprise organization address book, and sets the interface person in the engine structure design field as the person in charge of the standard, and automatically issues it to the other party.

[0065] S4043. Make compliance judgments on the test report results according to the requirements of standards and regulations, including automatic judgment of quantitative requirements and auxiliary judgment of qualitative requirements. At the same time, upload supporting materials for the judgment results and mark the risk level (red for high risk, yellow for medium risk, and green for low risk).

[0066] Among them, automatic determination of quantitative requirements: for quantitative requirements that limit certain aspects of automobile performance or characteristics through specific parameters, the system identifies the parameter types and determination conditions in the standards and regulations, reads the actual design parameters and / or test parameters of the corresponding vehicle model, determines whether the determination conditions are met and outputs the determination results.

[0067] Among them, for the qualitative requirements that define the performance or characteristics of automobiles in descriptive language, semantic analysis is used to assist in the judgment: First, the BERT model is used to perform semantic analysis on the standards and regulations to extract the keywords related to the qualitative requirements. Then, the test report results are matched with the keywords to determine the test information related to the standards and regulations. For example, Pandas is used to integrate the test report results and match them with the keywords to find out the information related to the regulatory requirements in the actual design and test parameters of the vehicle model. Finally, Scikit-learn is used to perform data analysis and calculate the consistency score between the test information and the qualitative requirements. For example, the classification model provided by Scikit-learn is used to determine whether the aforementioned test information related to the standards and regulations and the qualitative requirements belong to the same category, and the confidence of the category is used as the consistency score. Then, an analysis report containing the consistency between the actual parameters of the vehicle model and the qualitative requirements is generated through the Web framework and provided to the responsible person for reference. The responsible person confirms and submits all the judgment results based on the quantitative requirement judgment results and qualitative requirement analysis report output by the system, combined with the actual situation of the vehicle model, for review by experts.

[0068] S405. Summary of project compliance determination results: At the end of the project compliance determination process, all determination results (compliant, non-compliant, not involved, pending verification) are synchronized to the standard-related project analysis and vehicle model development compliance list. Non-compliant or pending verification items in the determination results will be synchronized to the non-compliant item tracking library.

[0069] Among them, the non-conforming item tracking library summarizes and displays the evaluation results of the feedback in the project compliance judgment as non-conforming and pending verification problems, and also includes subsequent rectification plans, such as rectification responsible persons, rectification measures, and expected rectification completion time; regulatory engineers can also manually add non-conforming and pending verification problems. After the non-conforming item tracking process is completed, the rectification status of the problem items involved in the process, such as completion status, actual rectification completion time, supporting materials, etc., will be synchronously updated to the non-conforming item tracking library; this module can perform module, operation button and data level authority control for different roles. The standard-related project analysis module is used to summarize and display the compliance status of each stage of each standard in each project through the compliance judgment process from the standard and regulatory dimensions; this module can perform module, operation button and data level authority control for different roles, which needs to be set through the system management unit-role management module. If the standards or clauses actually involved in the process exist in the R&D compliance list of a certain vehicle project, the judgment results of the corresponding standards or clauses will be synchronously updated to the list, that is, the R&D compliance list can realize the summary display of the compliance status of each node of each project through the compliance judgment process from the vehicle dimension.

[0070] S406. Correction process for non-conformance items: For all problem items that have not passed the compliance assessment, a correction process for non-conformance items is created to track the correction status until all correction measures for non-compliant data are completed and the final expert review is passed. The completion status and process data of each correction stage will be updated in real time in the non-conformance tracking module.

[0071] The data source of the non-conformance tracking library is the project line and regulatory line of the non-conformance determination process. It displays the standards / clauses corresponding to each problem item, related projects, problem types, responsible departments, rectification responsible persons, rectification plans for each stage (design rectification, verification rectification, final rectification), current rectification status, actual closure time, closure method, etc. in a list format. You can view the specific rectification process materials and the progress of the corresponding non-conformance rectification process.

[0072] For compliance issues that have not been identified in the project compliance determination process, it is allowed to directly create non-compliance problem items in the non-compliant item tracking module, which needs to be set up through the system management unit-role management module. It is allowed to directly and manually close non-compliant data in non-compliant items, which needs to be set up through the system management unit-role management module. After each stage of rectification is completed, the corresponding rectification status, supporting materials and completion time will be synchronously updated in the non-compliant item tracking module; for the problem items that have completed the final stage of rectification, an email or system reminder will be sent to the initiator of the rectification process for the problem item. During the entire non-compliant item rectification process, users can log in to the application system to check the process status at any time. When the process is initiated, after each process node is completed, it will automatically flow to the next process node and send real-time emails or system notifications to PCs and mobile phones to remind relevant personnel to ensure that the tasks are processed and submitted in a timely manner.

[0073] Optionally, the system provided in this embodiment also includes a standard life cycle management unit, which is used to manage the planning stage, enterprise standard formulation and revision stage, enterprise standard evaluation and supervision and inspection stage, and enterprise standard review and cancellation stage of the enterprise standard.

[0074] Specifically, the standard life cycle management unit realizes the compliance process control of the product project throughout its life cycle, realizes the automatic mapping of domestic and foreign standard and regulatory requirements with vehicle model projects, the automatic allocation of assessment tasks, and the automatic judgment of quantitative requirements, ensuring the timeliness, accuracy, and completeness of regulatory inspections and avoiding delays and omissions in the response process. By displaying the task allocation and specific review progress through a visual process, it can assist regulatory engineers, design engineers, professional experts, and administrative and technical department personnel to have a clear understanding of the overall compliance inspection progress and results of the product; the system performs hierarchical control of process data according to users of different roles to ensure data security.

[0075] like Figure 6 As shown in the figure, the standard life cycle management unit includes the following modules: enterprise standard project establishment / formulation and revision evaluation process, enterprise standard plan adjustment process, enterprise standard plan management ledger, enterprise standard formulation and revision process, enterprise standard supervision and inspection process, enterprise standard evaluation process, enterprise standard review process, and enterprise standard abolition process. At the same time, the standard online editing tool is called to quickly generate enterprise standard texts that meet the writing specifications in the enterprise standard formulation and revision process.

[0076] During the enterprise standard planning stage: the enterprise standard plan management ledger automatically aggregates plans from different sources into the plan management ledger in accordance with the standardized enterprise standard plan template, solving the problem of manual judgment, screening and sorting of various plan sources in the traditional enterprise standard management model, avoiding information legacy and verification time costs caused by inconsistent formats when collecting plans offline, and at the same time establishing a unified portal for publishing formulation and revision tasks, and automatically reminding delayed plans to ensure the timely implementation of enterprise standard formulation and revision tasks.

[0077] During the formulation and revision stage of enterprise standards: the main drafter can initiate the process through simple operations, and automatically obtain the enterprise standard plan and number, reducing human errors. Secondly, the system can automatically generate pre-made templates that meet the requirements according to the type of enterprise standard, including detailed settings such as chapter and clause numbers and font sizes, to ensure the uniformity and professionalism of document format specifications. Then, with the help of standard online editing tools, not only can documents be edited efficiently online, but also rich content such as tables and pictures can be directly inserted, and relevant information can be automatically obtained from other modules of the system, which improves the efficiency of enterprise standard text writing. Finally, the process also covers the solicitation of opinions, standardization review, multi-level approval and other links to ensure the quality and compliance of enterprise standard content, and can automatically archive the published enterprise standards, realizing closed-loop management of the entire formulation and revision process. This series of functions significantly improves the efficiency and accuracy of enterprise standard formulation and revision.

[0078] During the enterprise standard evaluation and supervision and inspection stage: In the approval of formulation and revision content, enterprise standard evaluation and supervision and inspection work, the workflow engine is used to assign responsibilities to corresponding persons in charge, thus realizing online process and visual management during the approval of enterprise standard-related business. At the same time, the formulation and revision progress, evaluation results, supervision and inspection results and rectification progress are synchronized to the corresponding modules in real time, making it easier to understand the execution of enterprise standard business.

[0079] During the review and cancellation stage of enterprise standards: the system automatically reviews the review date and automatically processes enterprise standards with different review results, improving review efficiency and ensuring the health and applicability of the enterprise standard system.

[0080] The following is a further detailed description of this unit. The specific implementation steps of this unit are as follows: S501. Execute the enterprise standard project establishment / formulation and revision assessment process, complete the annual or temporary formulation and revision plan approval and synchronize it to the enterprise standard plan management ledger as the main source for subsequent enterprise standard formulation and revision tasks.

[0081] The enterprise standard project establishment / formulation and revision assessment process collects, approves and publishes the annual standard formulation and revision plans and temporary plans within each department of the enterprise, covering the category, name, number, drafting department, plan start and end time and other plan information of the enterprise standards to be formulated or revised in the future. After the process is completed, the enterprise standard engineer does not need to manually screen and organize the plan lists applied by each department, but directly summarizes and stores them in the enterprise standard ledger.

[0082] S502. The enterprise standard plan management ledger synchronizes the enterprise standard formulation and revision plan information from different sources (manual addition, enterprise standard project establishment / formulation and revision process, enterprise standard evaluation process, enterprise standard review process) in real time, and manages and reviews each plan detail. Plans that need to be changed are completed through the enterprise standard plan adjustment process.

[0083] The enterprise standard plan management ledger manages the enterprise standard formulation and revision plan, and is also used to publish the enterprise standard formulation and revision tasks, and displays the status and related information of each enterprise standard formulation and revision process task in real time. For plans that have been postponed beyond the planned completion date, it supports red reminders for extensions and pushes them to the corresponding responsible persons. The enterprise standard plan comes from the following 4 methods: ① Manual addition, that is, adding a single plan or importing the enterprise standard formulation and revision plan in batches through the plan preparation template; ② Plans that have passed the enterprise standard project establishment / formulation and revision evaluation process; ③ Enterprise standards that are considered to need to be revised in the enterprise standard evaluation process; ④ Enterprise standards that are required to be revised in the enterprise standard review process.

[0084] The enterprise standard plan management ledger includes the plan master list and the detailed revision information page of each plan. The plan master list only displays the basic information of the revision plan, such as plan type, enterprise standard number, enterprise standard name, alternative enterprise standard number, drafter, drafting department, plan completion time, etc.; the detailed revision information page contains the basic information, related information, drafting information, etc. of each plan, such as plan source, plan completion status, other drafting departments, related process numbers, etc.

[0085] S503. Initiate the task of drafting and revising in the enterprise plan management ledger and complete the entry of the enterprise standard into the database: the drafter fills in the enterprise standard attribute information, selects the corresponding type of enterprise standard prefabricated template, and calls the standard text editing tool to write the content to generate the enterprise standard draft. After soliciting opinions, submitting for review, and submitting for approval, the enterprise standard text is officially released and automatically archived in the enterprise standard database after final approval; The process of making and revising enterprise standards is mainly divided into the following steps: 1. The main drafter initiates the process, selects the enterprise standard plan to be carried out in the ledger, enters the enterprise standard attribute information, and the enterprise standard number is automatically generated according to the corresponding rules.

[0086] 2. Generate standard prefabricated templates, provide various enterprise standard templates, and automatically generate text formats that meet the requirements according to the enterprise standard type corresponding to the plan, including chapter and clause numbers, fonts, font sizes, line spacing, chart numbers, headers and footers, etc. 3. After selecting the template, call the standard online editing function, support shortcut keys to insert numbers, charts, formulas, reference materials, set editing areas and other functions for online content writing and proofreading.

[0087] 4. Initiate solicitation of opinions, standardization review, and approval for the completed draft of the enterprise standard text, and pass the prepared / revised enterprise standard to other drafters or heads of departments to review the content of the enterprise standard and provide feedback on revisions.

[0088] 5. Summarize and analyze the revision opinions, revise and polish the content of the enterprise standard text once or multiple times to form the final draft for approval.

[0089] 6. After completing multi-level approval and joint signing of the modified enterprise standard text, the enterprise standard will be officially released and automatically archived. After the standard is released, it will be automatically included in the enterprise standard library for storage.

[0090] S504. Execute the enterprise standard evaluation process for newly stored enterprise standards: distribute the enterprise standard evaluation form to relevant standardization experts to score the quality of enterprise standard writing and fill in the review opinions, and complete the subsequent review. After the process is completed, the newly added or revised enterprise standards involved in the enterprise standard evaluation opinions will be included in the enterprise standard plan management ledger for subsequent enterprise standard formulation and revision work.

[0091] The enterprise standard evaluation process is used to evaluate and improve newly released enterprise standards. The process manages the enterprise standard evaluation form online and assigns it to relevant standardization experts for filling in. The experts score the quality of the enterprise standard writing and fill in the enterprise standard review opinions for subsequent review. After the process is completed, the newly added or revised enterprise standards involved in the enterprise standard evaluation opinions should be included in the enterprise standard plan management ledger for subsequent enterprise standard formulation and revision work.

[0092] S505. Execute the supervision and inspection process for the newly stored enterprise standards: Check the use of enterprise standards, compile and publish supervision and inspection plans, distribute and execute supervision and inspection tasks, generate supervision and inspection reports and rectification plans, rectify supervision and inspection issues, and track and close them. The supervision and inspection of enterprise standards is mainly divided into the following steps: 1. Initiate the enterprise standard supervision and inspection process, compile and publish the annual plan or single plan for standard supervision and inspection, confirm the supervision and inspection object, start and end time, inspection person in charge, etc., and notify relevant departments and personnel; 2. Allocate supervision and inspection tasks according to the entire enterprise standard or clause. According to the previous task push, automatically remind relevant professionals and personnel before the corresponding inspection completion time node; 3. Fill in and summarize the supervision and inspection report and the corresponding rectification plan. The supervision and inspection report includes enterprise standards / clause requirements, inspection time, implementation status, supporting materials, non-conformity description, etc. For the problem items that do not meet the enterprise standards or clause requirements, it is necessary to upload the rectification plan, including rectification measures, rectification responsible persons, planned rectification completion time, etc., and publish the supervision and inspection report and rectification plan through the system, notify relevant departments and personnel, and enter the inspection report and corresponding rectification plan into the supervision and inspection problem dashboard; The supervision and inspection problem dashboard uniformly displays and analyzes supervision and inspection problems and rectification plans, and displays enterprise standard supervision and inspection data in a list. For each problem, you can view the inspection report to understand the details of the supervision and inspection problems and the corresponding rectification plan. At the same time, the dashboard can display the rectification progress of each problem in real time to confirm whether the rectification has been completed or exceeded the planned rectification completion time.

[0093] The rectification process for supervision and inspection issues will push the issues in the supervision and inspection issue dashboard to the person responsible for rectification. The person responsible will receive a task reminder and submit the rectification status and supporting materials. If the rectification is not completed, he or she can apply for a rectification extension and explain the situation. After the rectification process for supervision and inspection issues is completed, the rectification status of the corresponding issue will be updated synchronously in the dashboard, published and pushed to relevant personnel to view the rectification results. The person in charge of inspection will view the rectification status reported by the person responsible for rectification in the dashboard. For issues that meet the rectification requirements, the dashboard can be automatically updated to a closed status. For rectification results that still do not meet the requirements, a rectification process can be initiated until the rectification is in place.

[0094] S506. Every day, the enterprise standards stored in the enterprise standard library are automatically judged whether the review date has arrived: for enterprise standards that are about to reach the review date, the enterprise standard review process is automatically initiated; for enterprise standards that are still valid after the review result, the next review time is updated; for enterprise standards that are revised after the review result, they are included in the enterprise standard plan management ledger; for enterprise standards that are invalid after the review result, they are automatically put into the enterprise standard cancellation process.

[0095] For the enterprise standards review process that has been stored in the enterprise standard library, the system regularly scans the enterprise standard library every day to see if the review date is about to arrive, triggers the automatic initiation process of the review conditions, and carries out the enterprise standard review task. After the enterprise standard review process is completed, the system issues a notification to handle the standard review results. The review results in the review process include continued validity, revision, and invalidation. If the review result is continued validity, the next review time of the current enterprise standard will be updated; if the review result is revised, it will be automatically included in the enterprise standard plan management ledger for enterprise standard preparation and revision plan work; if the review result is invalidation, the invalidation standard can be automatically initiated as an enterprise standard invalidation process, and the status of the enterprise standard will be converted to invalidation in the enterprise standard library.

[0096] Optionally, the system also includes a standard early warning module, which is used to remind the standards that are about to be implemented in the domestic and foreign standard and regulation libraries and the enterprise standard library. Each standard and regulation is filled in with the corresponding implementation time when it is entered into the library, including the standard implementation date, the implementation date of new models, and the implementation date of vehicles in production. For each standard, by setting the reminder time before the implementation time, it regularly scans whether domestic and foreign standards and enterprise standards need to be warned before their respective implementation dates. If the conditions are met, the corresponding standard information will be automatically pushed to the standard early warning module, and a system reminder or early warning email will be sent to the relevant responsible persons of the corresponding standards.

[0097] Optionally, the system also includes a system integration unit, which is used to achieve data sharing with business systems such as personnel management, project management, parts management, test management, and certification, and to achieve guidance of standards and regulations and their structured data on various business areas by continuing data exchange with upstream and downstream systems. Taking the unified certification system, project management, and parts management system as an example, data is exchanged with the system integration unit to provide data acquisition, push, query and other operations. By connecting with the project management system, vehicle model project information is obtained for real-time synchronization of vehicle model project library information, and the conformity assessment results of vehicle model projects are fed back to the project management system. By connecting with the parts management system, the parts tree structure and part number related information are obtained as the basis for binding standards and regulations and their clauses with parts. The unit can also integrate corporate emails through SMTP, and standard dynamics and process reminders in the system can remind users through emails.

[0098] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0099] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A structured processing and application system for automobile standards and regulations, characterized in that: include: Basic data control unit, which manages and maintains standards and regulations in the automotive field; A standard structured processing unit for structured processing and storage of the standard regulations; An input unit receives attribute information and a test report of a target vehicle model; The product project compliance control unit determines a matching first set of standards and regulations according to the attribute information; extracts keywords from the test process of the test report, and determines a second set of standards and regulations in the first set of standards and regulations through a keyword matching algorithm; determines a third set of standards and regulations in the second set of standards and regulations according to the matching of the step sequence and step content of the matching keywords; determines a fourth set of standards and regulations in the third set of standards and regulations by performing a similarity comparison on the test images; and determines whether the target vehicle model meets the standards according to the test requirements of each standard regulation in the fourth set of standards and regulations.

2. The system according to claim 1, characterized in that The attribute information includes vehicle model information, test item information, region information and time information; Determining a first set of matching standard regulations according to the attribute information includes: A first set of standard regulations that conforms to the vehicle model information and the test item information and is valid within the geographical information and the time information is determined according to the attribute information.

3. The system according to claim 1, characterized in that According to the matching of the step sequence and the step content of the matched keywords, a third standard regulation set is determined in the second standard regulation set, including: The following formula is used to calculate the matching degree M of the step sequence and step content of the matching keywords in the second standard regulation set and the test report; ； Among them, A and B are coefficients, and , i is the index of the keyword, N is the number of keywords, It is a key word in the standard regulations The steps are It is a key word in the testing process The steps are and is the matching keyword, express and The difference in the order of steps is express and The semantic distance of the step contents; The standard regulations for which M is greater than a set value are added to the third standard regulation set.

4. The system according to claim 1, characterized in that By performing similarity comparison on the test images, a fourth standard regulation set is determined from the third standard regulation set, including: Find test images around the steps where the matching keywords are located; An image similarity comparison algorithm and a keyword comparison algorithm of the test image annotation name are used to determine a fourth standard regulation set from the third standard regulation set.

5. The system according to claim 1, characterized in that Management and maintenance of standards and regulations in the automotive field, including: According to the standard status change rules, current time and the substitution relationship between standards, the status of each standard and regulation is determined, including: drafting, soliciting opinions, review, approval, publication, upcoming implementation, currently effective and abolished.

6. The system according to claim 1, characterized in that Structuring and storing the standards and regulations, including: For each standard regulation, confirm the scope of the split; Traverse each paragraph in the split range and mark the paragraphs that meet the clause identification rules with clause numbers; Identify paragraphs with the same clause number and merge them, sort the merged paragraphs by context order and clause level, and extract and split the content by clause; Text, table and image data are respectively used as the smallest storage unit for structured storage.

7. The system according to claim 1, characterized in that Management and maintenance of standards and regulations in the automotive field, including: Compare the differences between two similar or previous versions of standard regulations; The difference comparison includes: Preprocessing of two standard regulations; A deep learning model is used to perform semantic analysis on the pre-processed standard regulations to obtain two matching clauses; A distance algorithm was used to conduct a preliminary difference analysis between the two clauses; The Smith-Waterman algorithm and the diff algorithm are used to conduct a deep difference analysis of the two clauses; Quantify and identify preliminary difference analysis and in-depth difference analysis, and generate difference reports.

8. The system according to claim 1, characterized in that Determining whether the target vehicle model complies with the standard according to the test requirements of each standard regulation in the fourth standard regulation set includes: Through the project compliance determination process, each standard and regulation in the fourth standard and regulation set that the target vehicle model needs to meet is automatically assigned a responsible person, and an assessment task is created and executed; Determine whether the test report results meet the qualitative and quantitative requirements of each standard and regulation in the fourth standard and regulation set; Among them, the judgment methods for qualitative requirements include: The BERT model is used to perform semantic analysis on standards and regulations, and keywords related to qualitative requirements are extracted; Match test report results with keywords to identify test information related to standards and regulations; Scikit-learn was used to perform data analysis and calculate the consistency score between the test information and the qualitative requirements.

9. The system according to claim 1, characterized in that Also includes: The standard life cycle management unit manages the planning stage, formulation and revision stage, evaluation and supervision and inspection stage, and review and cancellation stage of enterprise standards.

Images