Structured template pre-checking method and device for vehicle-end signal cloud demand

Abstract

This invention relates to the field of vehicle network data processing technology, and discloses a method and apparatus for pre-verification of structured templates for vehicle-side signal cloud uploading requirements. The method includes: constructing a structured template and a signal standard knowledge base based on a standardized form template for vehicle-side signal cloud uploading from a car manufacturer and a CAN bus communication protocol template; performing vectorization transformation on the signal standard knowledge base to obtain a signal parameter semantic vector library; associating and binding the structured template and the signal parameter semantic vector library to form an intelligent structured template; and adapting and integrating the intelligent structured template with a corresponding form interaction plugin; triggering a verification process based on the completion status of the intelligent structured template; combining the structured template and the signal parameter semantic vector library to intelligently verify the filled content within the intelligent structured template; integrating the verification results; and generating and outputting corresponding verification guidance data. This invention improves the efficiency of the entire process and ensures stable vehicle-to-cloud data transmission.

Description

Technical Field

[0001] This invention relates to the field of vehicle network data processing technology, specifically to a structured template pre-verification method and device for vehicle-side signal cloud uploading requirements. Background Technology

[0002] With the rapid development of intelligent connected vehicle technology, various signals generated by the vehicle-side Controller Area Network (CAN) bus (such as battery temperature, vehicle speed, controller operating status, fault codes, etc.) need to be uploaded to the cloud via a Telematics Box (TBOX) or Gateway (GW) to provide core data support for vehicle data analysis, fault warning, operation and maintenance management, and performance optimization. Currently, automakers generally adopt a model of "collecting signals from fixed Excel templates + standardized protocol template specifications" to collect signal upload requests submitted by various controller departments. However, existing technical solutions have many core flaws:

[0003] (1) Static templates with no pre-verification: Excel templates are purely static files with only fixed fields for input. They cannot verify the input content in real time. Issues such as format errors and parameter deviations require manual line-by-line verification, which can easily flow to downstream processes and increase rectification costs.

[0004] (2) Inefficient manual processing: Reporters often use colloquial terms and non-standard abbreviations, requiring repeated manual confirmation, which is inefficient and prone to transmitting incorrect information due to misunderstandings.

[0005] (3) The verification rules are fixed and the maintenance cost is high: Traditional verification relies on hard coding. The rules cannot be flexibly adapted to scenarios such as protocol iteration and new model development. The code needs to be modified frequently, which makes maintenance difficult.

[0006] Based on the above problems, it is necessary to develop a new structured template pre-verification method and device for vehicle-side signal cloud uploading requirements. Summary of the Invention

[0007] In view of the shortcomings of the prior art, the purpose of this application is to provide a structured template pre-verification method and device for vehicle-side signal cloud transmission requirements, which replaces manual review with automated verification, avoids errors from the source, improves the efficiency of the whole process, and ensures the stability of vehicle-to-cloud data transmission.

[0008] In a first aspect, embodiments of this application provide a structured template pre-verification method for vehicle-side signal cloud uploading requirements, comprising the following steps:

[0009] Based on the standardized form template for cloud-based vehicle-end signals from car manufacturers and the CAN bus communication protocol template, a structured template and a signal standard knowledge base are constructed. The signal standard knowledge base is then vectorized to obtain a signal parameter semantic vector library.

[0010] The structured template and the signal parameter semantic vector library are associated and bound to form an intelligent structured template, and the intelligent structured template is adapted and integrated with the corresponding form interaction plugin.

[0011] Based on the completion status of the intelligent structured template, a verification process is triggered. Combining the structured template and the signal parameter semantic vector library, intelligent verification is performed on the filled content in the intelligent structured template.

[0012] The verification results are integrated to generate and output the corresponding verification guidance data.

[0013] In the above technical solution, this method transforms standardized forms and communication protocols into an intelligent verification carrier that is linked and bound together, overcoming the technical shortcomings of traditional manual verification that relies on human experience and has inconsistent verification standards. It realizes intelligent pre-management of vehicle-side signal cloud access requirement templates, avoiding problems such as non-standard form filling and signal parameters that do not meet communication protocol requirements in advance, effectively reducing the frequency of rework in subsequent cloud access and signal transmission links. Relying on form plugin adaptation and integration, it achieves seamless connection between template filling and verification processes, optimizing the overall process continuity of requirement submission and review.

[0014] One possible implementation involves constructing a structured template and a signal standard knowledge base, and then performing a vectorization transformation on the signal standard knowledge base to obtain a signal parameter semantic vector library, specifically:

[0015] The standardized form template's field layout, attribute types, and basic format requirements are analyzed to generate the structured template.

[0016] Extract the vehicle-side signal parameters from the CAN bus communication protocol template and construct the signal standard knowledge base associated with the intelligent structured template fields;

[0017] By calling the open-source semantic coding model interface, various parameters in the signal standard knowledge base are vectorized to obtain the signal parameter semantic vector library.

[0018] In the above technical solution, the form structure, signal standard, and semantic features are transformed into callable and matchable digital resources, realizing the deep binding between the template structure and the actual communication protocol of the vehicle. The parameter vectorization process is completed by relying on the open source semantic coding model, eliminating the need to develop a dedicated coding tool, reducing modeling costs, and laying the data foundation for the intelligent recognition of non-standard content in the future, adapting to the actual verification needs of vehicle signal cloud.

[0019] One possible implementation is that the structured template includes at least one of the following: field attributes, data type, format specifications, field relationships, and validation priority.

[0020] In the above technical solution, by clarifying and refining the core dimensions of the structured template, a comprehensive and multi-level verification basis system is constructed, covering the basic specifications of form filling to the deep logical connections. At the same time, the verification priority is set, which can realize the hierarchical and orderly execution of the verification process, prioritize the investigation of core required fields and key parameter errors, avoid invalid verification time consumption, improve the pertinence and efficiency of verification, and ensure that the verification rules fully cover the standardized requirements of car manufacturers to upload vehicle-side signals to the cloud.

[0021] One possible implementation is that the intelligent verification includes at least one of the following: basic format compliance verification, parameter consistency verification, mandatory item integrity verification, vehicle-side signal acquisition feasibility pre-verification, and cloud business rule compliance verification.

[0022] In the above technical solution, by integrating multi-dimensional verification content, a full-process pre-verification system is constructed, covering everything from basic form format to in-depth business feasibility. This system not only covers the basic compliance requirements for form filling, but also anticipates potential problems at the signal collection and cloud computing business levels, enabling proactive risk management and preventing project stagnation due to infeasible collection or non-compliance with rules after the requirement is submitted. This comprehensively ensures the rationality, feasibility, and standardization of vehicle-side signal cloud computing requirements.

[0023] One possible implementation method is as follows: The parameter consistency check specifically involves:

[0024] The signal parameters filled in by the intelligent structured template are vectorized and encoded through the open-source semantic coding model interface to obtain a temporary semantic vector. The temporary semantic vector is then matched with the standard parameter semantic vector in the signal parameter semantic vector library for similarity. If the similarity is within a preset threshold range, suggestions for modifying the filled content are provided.

[0025] In the above technical solution, in response to common problems such as synonymous abbreviations, colloquial terms, and non-standard expressions in the process of vehicle-side signal reporting, the solution achieves accurate matching and standardized transformation of fuzzy semantics through similarity matching. This breaks the limitation of traditional verification that can only match completely identical text, greatly reduces the problems of misjudgment and omission caused by non-standard expressions, improves the accuracy and fault tolerance of parameter consistency verification, reduces the workload of manual intervention and correction, and adapts to the expression differences in reporting needs of different personnel.

[0026] In one possible implementation, the feasibility pre-verification of vehicle-side signal acquisition and the compliance verification of cloud business rules are both executed collaboratively through preset rule prompts and open-source large model interfaces, and the two verifications are completed synchronously through a single interface call.

[0027] In the above technical solution, special verification is completed by combining prompt word engineering with open source large model, eliminating the need to build a dedicated model, reducing R&D and deployment costs. At the same time, the two related verifications are merged into a single interface call, reducing the frequency of AI model interface calls, shortening the overall verification time, improving the running efficiency of the verification process, avoiding resource waste caused by repeated calls, and balancing verification accuracy and execution efficiency.

[0028] One possible implementation further includes: displaying the verification guidance data, which includes guidance content on marking error information and correction basis, and re-triggering verification after completing the correction of the marked error information until all corresponding verification dimensions pass.

[0029] In the above technical solution, error information and corresponding correction guidance are displayed visually, enabling the person submitting the request to quickly locate the problem and clarify the direction of correction without repeated communication and consultation, thus reducing the difficulty of correction. At the same time, a cyclical verification mechanism is set up to ensure that all error items are thoroughly rectified, realize closed-loop verification and control of form templates, and finally output the vehicle-side signal cloud access requirements that fully comply with the standards, ensuring the smooth progress of the entire subsequent signal cloud access process.

[0030] Secondly, this application provides a structured template pre-verification device for vehicle-side signal cloud uploading requirements, comprising:

[0031] The knowledge base and vector library construction module is used to construct a structured template and a signal standard knowledge base based on the standardized form template for cloud-based vehicle-end signals of car manufacturers and the CAN bus communication protocol template, and to perform vectorization encoding on the signal standard knowledge base to obtain a signal parameter semantic vector library.

[0032] The intelligent template integration module is used to associate and bind the structured template and the signal parameter semantic vector library to form an intelligent structured template, and to adapt and integrate the intelligent structured template with the corresponding form interaction plugin to realize the standardized filling interaction of vehicle-side signal cloud access requirements;

[0033] The intelligent verification execution module is used to trigger the verification process based on the completion status of the intelligent structured template, and to perform intelligent verification on the user-filled content in the intelligent structured template by combining the pre-built structured template and the signal parameter semantic vector library.

[0034] The verification result processing module is used to integrate and classify the various verification results output by the intelligent verification execution module, and generate and output the corresponding verification guidance data.

[0035] In the above technical solution, the device realizes the full-process function of AI modeling, intelligent verification, and result feedback through modular division of labor and cooperation. The responsibilities between modules are clear and the linkage is close. It can stably realize the pre-intelligent verification of the vehicle signal cloud demand template. Compared with traditional manual verification devices, it has the advantages of high automation, unified verification standards and strong adaptability. It can be directly integrated into the vehicle enterprise's vehicle network management system, quickly implemented and applied, and effectively improve the efficiency of demand review.

[0036] One possible implementation, the knowledge base and vector library construction module specifically includes:

[0037] The structure graph construction unit is used to parse the field layout, attribute types and basic format requirements of the standardized form template and generate the structured template.

[0038] The knowledge base construction unit is used to extract vehicle-side signal parameters from the CAN bus communication protocol template and construct the signal standard knowledge base associated with the intelligent structured template fields.

[0039] The semantic vector library generation unit is used to call the open-source semantic coding model interface to perform vectorization transformation on the parameters in the signal standard knowledge base to obtain the signal parameter semantic vector library.

[0040] In the above technical solution, by refining the internal units of the AI ​​modeling module, the modeling process can be finely divided and controlled. Each unit corresponds to a dedicated modeling task, ensuring the construction quality of structured templates, signal standard knowledge bases, and semantic vector libraries. The units cooperate with each other to form a complete modeling system, further enhancing the adaptability of templates to vehicle-side signal standards.

[0041] One possible implementation also includes a display and loop verification triggering module, used to display the verification guidance data, which includes guidance content on marking error information and correction basis, and to re-trigger verification after the marked error information is corrected, until all corresponding verification dimensions pass.

[0042] In the above technical solution, by adding a dedicated display and cyclic verification module, the closed-loop management function of the entire device is improved. The visual display interface enhances the user experience, while the cyclic verification mechanism completely eliminates residual errors, ensuring that the final submitted vehicle signal cloud upload requirement template fully complies with the vehicle manufacturer's standards and CAN bus communication protocol requirements. No manual secondary verification is required, further reducing labor costs and improving the automation and standardization of the entire verification process. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the background art, the accompanying drawings used in the embodiments of this application will be described below.

[0044] Figure 1 This is a flowchart of a structured template pre-verification method for vehicle-side signal cloud uploading requirements disclosed in an embodiment of this application;

[0045] Figure 2 This is one of the block diagrams of the structured template pre-verification device for vehicle-side signal cloud uploading requirements disclosed in the embodiments of this application;

[0046] Figure 3 This is a block diagram of the knowledge base and vector library construction module disclosed in an embodiment of this application;

[0047] Figure 4 This is a block diagram of the intelligent verification execution module disclosed in an embodiment of this application;

[0048] Figure 5 This is a second block diagram of the structured template pre-verification device for vehicle-side signal cloud uploading requirements disclosed in an embodiment of this application;

[0049] Figure 6 This is the third block diagram of the structured template pre-verification device for vehicle-side signal cloud uploading requirements disclosed in the embodiments of this application;

[0050] Explanation of reference numerals in the attached figures:

[0051] 1-Knowledge base and vector library construction module, 11-Structure graph construction unit, 12-Knowledge base construction unit, 13-Semantic vector library generation unit, 2-Intelligent template integration module, 3-Intelligent verification execution module, 31-Open source Embedding model interface calling unit, 32-Prompt+DeepSeek large model interface calling unit, 33-AI verification engine unit, 4-Verification result processing module, 5-Display and loop verification triggering module, 6-Data storage module. Detailed Implementation

[0052] The embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The detailed description of the following embodiments and the accompanying drawings are used to illustrate the principles of this application by way of example, but should not be used to limit the scope of this application, that is, this application is not limited to the described embodiments.

[0053] Please see Figure 1 , Figure 1 This is a flowchart illustrating a structured template pre-verification method for vehicle-side signal cloud uploading requirements disclosed in an embodiment of this application. The method includes the following steps:

[0054] Based on the standardized form templates for cloud-based vehicle-end signals from automotive manufacturers and the CAN bus communication protocol template, a structured template and a signal standard knowledge base are constructed. The signal standard knowledge base is then vectorized to obtain a semantic vector library of signal parameters.

[0055] The structured template and the signal parameter semantic vector library are linked and bound to form an intelligent structured template, and the intelligent structured template is adapted and integrated with the corresponding form interaction plugin.

[0056] Based on the intelligent structured template completion status trigger verification process, combined with the structured template and signal parameter semantic vector library, the content filled in the intelligent structured template is intelligently verified.

[0057] The verification results are integrated to generate and output corresponding verification guidance data (such as error information and guidance content for correction).

[0058] This application utilizes AI-based structured modeling to transform standardized forms and communication protocols into an intelligent, interconnected verification mechanism, overcoming the technical shortcomings of traditional manual verification which relies on human experience and suffers from inconsistent verification standards. It achieves intelligent pre-emptive management of vehicle-side signal cloud upload requirement templates, proactively avoiding issues such as non-standard form completion and signal parameters not meeting communication protocol requirements. This effectively reduces the frequency of rework in subsequent cloud integration and signal transmission stages. Through form plugin adaptation and integration, it achieves seamless connection between template completion and verification processes, optimizing the overall flow of requirement submission and review.

[0059] In this embodiment of the application, a structured template and a signal standard knowledge base are constructed, and the signal standard knowledge base is vectorized to obtain a signal parameter semantic vector library, specifically as follows:

[0060] This paper analyzes the field layout, attribute types, and basic formatting requirements of standardized form templates to generate a structured template. For example, using an Excel spreadsheet as an example, it analyzes the field layout, attribute types (e.g., required / optional), and basic formatting requirements (e.g., text / number / enumeration) of a fixed Excel template to generate a visual structured template (i.e., a comma-separated list of structured template rules). This clarifies the relationships and validation priorities of each field, providing a basis for subsequent format validation. In this application, " / " represents an alternative relationship.

[0061] Extract vehicle-side signal parameters (such as signal ID, signal name, sampling frequency, vehicle model compatibility range, signal start position, and transceiver controller) from the CAN bus communication protocol template, and construct a signal standard knowledge base associated with intelligent structured template fields.

[0062] By calling open-source semantic coding model interfaces (such as open-source Embedding model interfaces), various parameters (such as signal names, transceiver controllers, and other text information) in the signal standard knowledge base are vectorized to generate a signal parameter semantic vector library.

[0063] Then, the structured template is bound to the signal parameter semantic vector library to generate an intelligent structured template, which is then packaged as an Excel plugin and integrated into the Excel application, allowing users to trigger verification after filling in the information. Users can use the Excel spreadsheet with the integrated plugin to fill in the relevant information for vehicle-side signal cloud uploading requirements according to the requirements of the intelligent structured template.

[0064] This application transforms form structures, signal standards, and semantic features into callable and matchable digital resources, achieving deep binding between the template structure and the actual communication protocol on the vehicle side. It relies on an open-source semantic coding model to complete parameter vectorization processing, eliminating the need to develop dedicated coding tools, reducing modeling costs, and laying a data foundation for the intelligent recognition of non-standardized content, thus adapting to the actual verification needs of vehicle-side signals in the cloud.

[0065] In this embodiment, the structured template includes at least one of the following: field attributes, data type, format specifications, field relationships, and validation priority. This application constructs a comprehensive, multi-layered validation framework by clearly defining and refining the core dimensions of the structured template. This framework covers everything from basic form filling specifications to deep logical relationships. Simultaneously, by setting validation priorities, it enables hierarchical and orderly execution of the validation process, prioritizing the detection of errors in core required fields and key parameters, avoiding time-consuming invalid validations, improving the targeting and efficiency of validation, and ensuring that validation rules fully cover the standardized requirements for vehicle-side signal uploading to the cloud.

[0066] In this embodiment of the application, intelligent verification includes at least one of the following: basic format compliance verification, parameter consistency verification, mandatory item integrity verification, vehicle-side signal acquisition feasibility pre-verification, and cloud business rule compliance verification. Wherein:

[0067] Basic format validation: Based on the structured template, the format, capitalization (automatically unified according to the car manufacturer's standards), length, and data type of the filled content are validated to ensure they meet the requirements. If they do not meet the requirements, the error location and type are directly marked.

[0068] Parameter consistency verification: The signal parameters (unstructured content entered by the user, such as synonyms, abbreviations, and colloquialisms) entered through the intelligent structured template are vectorized using an open-source semantic encoding model interface (e.g., the open-source Embedding model interface) to obtain temporary semantic vectors. These temporary semantic vectors are then matched with standard parameter semantic vectors in the signal parameter semantic vector library for similarity. If the similarity is within a preset threshold range, a modification suggestion is provided. For example, the preset threshold range is greater than or equal to 0.9 and less than 1. For instance, if the similarity between "battery temperature," "cell temperature," and the standard name "battery temperature" is within the preset threshold range, when the user enters "battery temperature" or "cell temperature," the device will suggest changing "battery temperature" or "cell temperature" to "battery temperature." Similarly, if the similarity between "vehicle gateway" and the standard name "TBOX" is within the preset threshold range, when the user enters "vehicle gateway," the device will suggest changing it to "TBOX."

[0069] Required field validation: Automatically validates whether all required fields in the intelligent structured template are filled in completely. If any field is missing, it will be marked as missing to avoid missing core information.

[0070] Feasibility pre-verification of data collection: This is achieved through a prompt word and the DeepSeek open-source big model interface. The core rules for the feasibility of vehicle-side signal collection are written into a fixed prompt word in advance. The core parameters such as the signal name, receiver controller, sampling frequency, and vehicle type filled in by the user are synchronously transmitted to the DeepSeek open-source big model along with the fixed prompt word. The DeepSeek open-source big model directly determines whether the signal meets the gateway collection conditions and whether it supports the subsequent cloud process, quickly identifying problems such as limited collection and parameters that do not meet the collection requirements.

[0071] Business rule verification: This is also accomplished using the Prompt+DeepSeek open-source large model interface. The core business rules for vehicle-side signals to the cloud (such as only signals from controllers including Power Supply Management Controller (SVDC) and Telematics Box (BOX) can be uploaded to the cloud) are written into the prompt words. The DeepSeek open-source large model is then called to complete the business rule compliance determination and output the rule verification results.

[0072] This application integrates multi-dimensional verification content to construct a full-process pre-verification system covering everything from basic form format to in-depth business feasibility. It not only covers the basic compliance requirements for form filling, but also anticipates potential problems at the signal collection and cloud computing business levels, enabling proactive risk control and preventing projects from stalling due to infeasible collection or non-compliance with rules after the requirement is submitted. It comprehensively ensures the rationality, feasibility, and standardization of vehicle-side signal cloud computing requirements.

[0073] In this embodiment, the feasibility pre-verification of vehicle-side signal acquisition and the compliance verification of cloud business rules are both executed collaboratively through preset rule prompts and open-source large model interfaces, and the two verifications are completed synchronously through a single interface call.

[0074] This application uses prompt word engineering combined with open-source large models to complete special verification, eliminating the need to build a dedicated model, reducing R&D and deployment costs. At the same time, it merges two related verifications into a single interface call, reducing the frequency of AI model interface calls, shortening the overall verification time, improving the running efficiency of the verification process, avoiding resource waste caused by repeated calls, and balancing verification accuracy and execution efficiency.

[0075] In this embodiment, a structured template pre-verification method for vehicle-side signal cloud uploading requirements further includes: displaying the generated error information (including error type, error location, and specific reason) and the guidance content for correction (e.g., "Signal name is not standardized: it is recommended to correct it to the standard expression 'battery temperature'", "Controller mismatch: only SVDC / TBOX controllers support this signal uploading to the cloud", "Collection feasibility fails: the current sampling frequency exceeds the gateway's allowed range, it is recommended to adjust it to 1-10Hz"). For the matching results of unstructured content, the correction logic is displayed intuitively (e.g., "Currently filled in 'battery temperature', matches the standard signal name 'battery temperature', according to the standard parameters of the CAN bus communication protocol"), allowing users to clearly understand the reasons for correction. All error items and correction guidance are centrally displayed in the Excel plugin interface. Users can manually complete the content correction according to the guidance. After correction, the verification function can be retried until all verification dimensions pass.

[0076] This application uses a visual display of error information and corresponding correction guidance to enable those submitting requests to quickly locate problems and clarify the direction of correction without repeated communication and consultation, thus reducing the difficulty of correction. At the same time, a cyclical verification mechanism is set up to ensure that all errors are thoroughly rectified, realizing closed-loop verification and control of the form template, and finally outputting vehicle-side signal cloud access requirements that fully comply with the standards, ensuring the smooth progress of the entire subsequent signal cloud access process.

[0077] Please see Figure 2 , Figure 2This is one of the block diagrams of a structured template pre-verification device for vehicle-side signal cloud access requirements disclosed in this application. The structured template pre-verification device for vehicle-side signal cloud access requirements is used to implement a structured template pre-verification method for vehicle-side signal cloud access requirements. It adopts an architecture of Excel plugin + open-source AI interface call, enabling lightweight deployment and compatibility with Office / WPS Excel applications on Windows / macOS systems. It provides core verification and data processing functions, with all components working collaboratively to complete the entire process of template modeling, centralized verification, and result feedback. Specifically, the structured template pre-verification device for vehicle-side signal cloud access requirements includes a knowledge base and vector library construction module 1, an intelligent template integration module 2, an intelligent verification execution module 3, and a verification result processing module 4. The knowledge base and vector library construction module 1 is used to construct a structured template and a signal standard knowledge base based on the standardized form template for vehicle-side signal cloud access from automotive manufacturers and the CAN bus communication protocol template. The signal standard knowledge base is then vectorized to obtain a signal parameter semantic vector library. The intelligent template integration module 2 is used to associate and bind structured templates and signal parameter semantic vector libraries to form intelligent structured templates. It then adapts and integrates these intelligent structured templates with corresponding form interaction plugins, enabling standardized filling and interaction for vehicle-side signal cloud upload requirements. The intelligent verification execution module 3 triggers the verification process based on the completion status of the intelligent structured template. Combining the pre-built structured template and signal parameter semantic vector library, it intelligently verifies the user-filled content within the intelligent structured template. The verification result processing module 4 integrates and categorizes the various verification results output by the intelligent verification execution module 3, generating and outputting corresponding verification guidance data. This device, through modular division of labor and collaboration, achieves full-process functionality encompassing AI modeling, intelligent verification, and result feedback. The modules have clear responsibilities and close linkage, enabling stable pre-process intelligent verification of vehicle-side signal cloud upload requirement templates. Compared to traditional manual verification devices, it boasts advantages such as high automation, unified verification standards, and strong adaptability. It can be directly integrated into vehicle manufacturers' vehicle networking management systems for rapid application deployment and effectively improve requirement review efficiency.

[0078] Please see Figure 3 , Figure 3This is a block diagram of the knowledge base and vector library construction module disclosed in this application. The knowledge base and vector library construction module 1 specifically includes a structure graph construction unit 11, a knowledge base construction unit 12, and a semantic vector library generation unit 13. The structure graph construction unit 11 is used to parse the field layout, attribute types, and basic format requirements of standardized form templates to generate structured templates. The knowledge base construction unit 12 is used to extract vehicle-side signal parameters from the CAN bus communication protocol template and construct a signal standard knowledge base associated with the fields of the intelligent structured template. The semantic vector library generation unit 13 is used to call the open-source semantic coding model interface to perform vectorization conversion on various parameters in the signal standard knowledge base, generating a signal parameter semantic vector library. This application, by refining the internal units of the knowledge base and vector library construction module 1, achieves refined decomposition and control of the modeling process. Each unit corresponds to a dedicated modeling task, ensuring the construction quality of the structured template, signal standard knowledge base, and signal parameter semantic vector library. The units cooperate with each other to form a complete modeling system, further strengthening the adaptability of the template to the vehicle-side signal standard.

[0079] After constructing the structured template and signal parameter semantic vector library, the structured template and signal parameter semantic vector library are bound together to form an intelligent structured template. This intelligent structured template is then adapted and integrated with form interaction plugins (such as Excel plugins). The Excel plugin serves as the interaction point between hardware and software, embedded in the Excel application, providing a visual interface and supporting functions such as user template filling, validation triggering, and error result display. It aligns with existing office habits and requires no additional learning curve.

[0080] Please see Figure 4 , Figure 4 This is a block diagram of the intelligent verification execution module disclosed in this application. The intelligent verification execution module 3 includes an open-source Embedding model interface calling unit 31, a Prompt+DeepSeek large model interface calling unit 32, and an AI verification engine unit 33. The open-source Embedding model interface calling unit 31 and the Prompt+DeepSeek large model interface calling unit 32 respectively implement semantic encoding, vector matching, and standardized transformation of unstructured content. The Prompt+DeepSeek large model interface calling unit 32 is used to implement the Prompt solidification, parameter input, and dual rule compliance determination of the vehicle-side signal cloud business rules and collection feasibility rules. The AI ​​verification engine unit 33, as the core component of the device, runs in series with the modules, receives verification instructions, and performs multi-dimensional intelligent verification.

[0081] In this embodiment, the verification result processing module 4 is used to connect the AI ​​verification engine and the Excel plugin, and to integrate the full verification results. It marks the error type, location and cause for each error, and generates targeted correction guidance suggestions, covering various issues such as format errors, semantic non-standardization, business rule inconsistencies, and failure to pass data collection feasibility tests, providing clear basis for users to manually correct errors.

[0082] like Figure 5 As shown, Figure 5 This is a second block diagram of a structured template pre-verification device for vehicle-side signal cloud uploading requirements disclosed in an embodiment of this application. The structured template pre-verification device for vehicle-side signal cloud uploading requirements further includes a display and loop verification trigger module 5, used to display the generated error information and guidance content for correction, and to re-trigger verification after correcting the error information, until all corresponding verification dimensions pass, thus achieving closed-loop verification.

[0083] Please see Figure 6 , Figure 6 This is the third block diagram of a structured template pre-verification device for vehicle-side signal cloud uploading requirements disclosed in this application. The structured template pre-verification device for vehicle-side signal cloud uploading requirements also includes a data storage module 6, which employs a lightweight storage scheme to store data such as a signal standard knowledge base, a signal parameter semantic vector library, structured templates (i.e., a list of structured template rules), user-filled data, and verification results, enabling real-time data synchronization and interaction and avoiding data silos.

[0084] The following example illustrates the cloud-based signal collection requirements of a new energy vehicle manufacturer, which involves multiple vehicle models and controllers. The manufacturer's existing Excel requirement template includes core fields such as signal ID, signal name, controller model, sampling frequency, vehicle model compatibility, signal period, signal start position, and offset. Standardized protocol templates cover standard signals from various controller types, including Battery Management System (BMS), Vehicle Control Unit (VCU), and Motor Control Unit (MCU). The solution addresses issues such as user errors, inefficient processing of unstructured content, difficulty in rule inheritance, and cumbersome requirement submission processes. The device described in this application will be used for implementation, with the specific steps as follows:

[0085] Step 1: Obtain the fixed vehicle-side signal cloud upload Excel template from the automaker, define the cell layout, set required / optional field attributes, organize the core parameters of the CAN bus communication protocol, and compile a list of structured template rules and a list of standard signal parameters. The specific rules are as follows:

[0086] Column A is for signal ID, Column B is for signal name, Column C is for receiver controller, Column D is for sampling frequency, Column E is for vehicle model compatibility, Column F is for the name of the person submitting the request, Column G is for the department of the person submitting the request, Column H is for the reason for the request, and Column I is for remarks. Column I is optional for remarks, all others are required. Requirement-related fields are used to clarify the attribution and basis of the request, facilitating subsequent requirement traceability and internal circulation. The remarks field is used to supplement special instructions; it can be left blank if there are no special requirements. The core filling rules for each field are clearly defined in advance: signal ID is a combination of uppercase letters and integers, and cannot be purely numeric; the receiver controller must include SVDC and TBOX; the sampling frequency is only allowed to have three fixed values: 1s, 10s, and 60s. At the same time, the core parameters of the CAN bus communication protocol are compiled into two comma-separated text lists:

[0087] ① List of rules for structured templates (for format validation):

[0088] Cell location, field name, whether it is required, data type, format requirements;

[0089] A1, Signal ID, is, text, a combination of uppercase English letters and integers, 8-12 characters;

[0090] B1, Signal Name, Yes, Text, Uppercase Canonical Name, within 20 characters;

[0091] C1, Receive Controller, Yes, Multiple controllers are allowed, must include either SVDC or TBOX type, and multiple controllers should be separated by semicolons;

[0092] D1, sampling frequency, is enumerated, limited to three fixed values: 1s, 10s, and 60s.

[0093] E1, Vehicle Model Compatibility Range, Yes, Text, Enter the corresponding vehicle model series, within 30 characters;

[0094] F1, Requester's Name, Yes, Text, Chinese Name, within 10 characters;

[0095] G1, Requesting Department, Yes, Text, Fill in the full standard name of the department, no fixed enumeration limit, within 30 characters;

[0096] H1, Reason for Request, Yes, Text, Briefly explain the purpose of cloud migration, within 50 characters;

[0097] I1, Remarks, No, Text, Additional Special Notes, within 50 characters, can be left blank.

[0098] ② And the signal standard knowledge base (i.e., a list of standard signal parameters used to generate a semantic vector library):

[0099] BAT001, battery temperature, SVDC, 10s, pure electric SUV;

[0100] VCU002, vehicle speed, TBOX, 1s, all models;

[0101] BMS003, motor speed, BMS, 60s, pure electric vehicle;

[0102] BAT004, Remaining Battery Power, SVDC, 10s, Pure Electric SUV.

[0103] Step 2: Import the standard signal parameter list into the open-source Embedding model interface (such as the BGE-small open-source vector model) in batches, vectorize the signal names and receiver controller text data, generate signal parameter semantic vectors and store them in the plugin's local lightweight database to obtain the signal parameter semantic vector library.

[0104] Step 3: Write fixed prompts, synchronously integrate data collection feasibility and cloud business rules, and solidify them into the large model interface call unit of the Excel plugin. Subsequent verification can directly call them without repeated configuration.

[0105] For example, you are a vehicle-side signal cloud verification specialist at a car company. You strictly follow the following rules to make compliance judgments on the content filled in by users, and only output the judgment result and the specific reason for failure, in concise and straightforward language.

[0106] Business rules: Only signals from controllers containing either SVDC or TBOX are allowed to be uploaded to the cloud; otherwise, the application will be rejected.

[0107] Feasibility rules for data collection: Both of the following requirements must be met simultaneously; failure to meet either requirement will result in disqualification.

[0108] ① The signal ID must be a combination of uppercase English letters and an integer; pure numbers are strictly prohibited.

[0109] ② The sampling frequency is limited to three fixed values: 1s, 10s, and 60s. Other values ​​cannot be collected.

[0110] Input parameters: signal ID, signal name, receiver controller, sampling frequency.

[0111] The prompt is embedded into the DeepSeek interface call unit of the Excel plugin, so that subsequent verifications can be directly called without repeated configuration.

[0112] Step 4: Bind the structured template and the signal parameter semantic vector library together to generate an intelligent structured template, embedding it into an Excel plugin. After modeling is complete, users can manually trigger validation after filling in the information. The filling process has no real-time AI prompts or formatting guidance, completely conforming to existing office filling habits.

[0113] Step 5: Users open the Excel template equipped with the verification plugin, fill in the cloud migration requirement information according to the column headings and field rules, and fill in the information independently without AI intervention. Required fields must be filled in completely and in compliance with regulations, while remarks can be filled in or left blank.

[0114] Example input: A1: 1001, B1: Battery temperature, C1: BMS, D1: 15s, E1: Pure electric SUV, F1: Zhang San, G1: Intelligent connected vehicle R&D department, H1: Vehicle remote status monitoring, I1: Winter working condition special test; After completing the input, click the Start Verification button on the plugin interface to manually trigger the full process verification. The plugin captures all the input data in real time and sends it to the AI ​​verification engine.

[0115] Step 6: The validation engine performs multi-dimensional validations sequentially, without any manual intervention. It relies solely on open-source interfaces and preset lists for execution. The remarks fields do not participate in any validation logic. The specific process is as follows:

[0116] 6.1. Basic Format + Mandatory Item Verification: Check each of the eight mandatory fields against the structured template rule list to ensure there are no gaps. Pay special attention to verifying that the signal ID is a combination of English letters and integers, the receiver controller is filled in correctly, the sampling frequency is at a compliant level, and the number of characters for the requester's department does not exceed 30 characters. This step detected two format errors: the signal ID being purely numeric and the sampling frequency being at a non-compliant level.

[0117] 6.2. Parameter Consistency Verification: The open-source Embedding interface is called to semantically encode the "battery temperature" entered by the user to generate a temporary semantic vector. This vector is then compared with the standard signal name in the signal parameter semantic vector library. If the similarity meets the standard, the expression is judged to be non-standard, marked as incorrect, and a standard name correction suggestion is generated.

[0118] 6.3. Business + Data Acquisition Feasibility Dual Verification: Extract the signal ID, signal name, receiver controller, and sampling frequency filled in by the user, concatenate them with the preset prompt, and call the DeepSeek open-source large model interface. The model synchronously completes the dual judgment according to the fixed rules and outputs the reason and result of failure.

[0119] 6.4. Verification Result Integration: The engine summarizes all error messages, removes data from the remarks field, forms a complete error list, and transmits it to the result feedback module.

[0120] Step 7. The validation engine displays all errors in a pop-up window on the right side of the Excel add-in. Each error is marked with its location, type, cause, and suggested correction. Example display:

[0121] Cell A1 (Signal ID): The format is not standard. The signal ID should be changed to a combination of uppercase letters and an integer.

[0122] Cell B1 (Signal Name): The description is not standardized and should be corrected to the standard name "Battery Temperature".

[0123] Cell C1 (Receiver Controller): Business rules do not comply, does not contain SVDC or TBOX, not allowed to go to the cloud;

[0124] Cell D1 (Sampling Frequency): Data acquisition is not feasible; only three settings are allowed: 1s / 10s / 60s. Adjustment is recommended.

[0125] The user manually corrects the content based on the prompts. Example of compliant data after correction:

[0126] A1=BAT001, B1=Battery Temperature, C1=SVDC, D1=10s, E1=Pure Electric SUV, F1=Zhang San, G1=Intelligent Connected Vehicle R&D Department, H1=Vehicle Remote Status Monitoring, I1=Winter Working Condition Special Test.

[0127] After making the corrections, click "Start Verification" again to re-trigger the full verification process until all verifications pass.

[0128] Compared with the prior art, this invention has the following significant advantages: it aligns with the actual business needs of automakers, is highly practical, and has low implementation costs:

[0129] 1. Eliminate errors at the source: Through real-time pre-entry verification and AI-powered intelligent error correction, more than 85% of formatting errors, parameter deviations, and protocol mismatches can be eliminated, preventing errors from flowing to downstream processes and significantly reducing the cost of manual verification and rectification; at the same time, AI guidance avoids common filling pitfalls and further improves the accuracy of filling.

[0130] 2. Efficiently solve the problem of unstructured content processing: Relying on the open-source Embedding model interface, it can accurately identify and standardize the conversion of synonyms, abbreviations and colloquialisms, without the need for manual confirmation one by one, improving the efficiency of requirement collection by more than 90% and avoiding the transmission of errors due to misunderstandings.

[0131] 3. Lightweight deployment, adaptable to existing office scenarios: Using an Excel plugin as the carrier, it does not require modification of the car company's existing business processes, fits the existing Excel work habits of employees, has no additional learning cost, and is compatible with Windows / macOS systems and Office / WPS. It has low deployment cost and fast deployment speed, and can be quickly scaled up to various controller departments; at the same time, the plugin supports network connectivity and realizes AI guidance, further improving ease of use.

[0132] 4. Flexible rule verification and adaptation without code development: The Prompt+DeepSeek open-source large model interface enables dual verification of business rules and data collection feasibility. Automakers can flexibly modify the rule requirements in the Prompt prompts according to their actual needs, without the need for technical personnel to frequently modify the code. This adapts to different vehicle models and controllers, enabling signal cloud uploading and data collection rule adjustments, thus reducing tool adaptation costs.

[0133] 5. Open source model interface calls reduce technical R&D costs: The core capabilities are realized entirely through open source AI model (Embedding model, DeepSeek large model) interfaces, eliminating the need for enterprises to independently develop and train AI models, greatly reducing technical R&D and maintenance costs, and adapting to the lightweight usage needs of SMEs and various business departments of car companies.

[0134] The examples are not limited to those described above. Those skilled in the art can make modifications or alterations based on the above description, and all such modifications and alterations should fall within the scope of protection of the appended claims. Those skilled in the art can understand that implementing all or part of the processes of the above embodiments and making equivalent changes according to the claims of this application still fall within the scope of this application.

Claims

1. A structured template pre-verification method for vehicle-side signal cloud uploading requirements, characterized in that, Includes the following steps: Based on the standardized form template for cloud-based vehicle-end signals from car manufacturers and the CAN bus communication protocol template, a structured template and a signal standard knowledge base are constructed. The signal standard knowledge base is then vectorized to obtain a signal parameter semantic vector library. The structured template and the signal parameter semantic vector library are associated and bound to form an intelligent structured template, and the intelligent structured template is adapted and integrated with the corresponding form interaction plugin. Based on the completion status of the intelligent structured template, a verification process is triggered. Combining the structured template and the signal parameter semantic vector library, intelligent verification is performed on the filled content in the intelligent structured template. The verification results are integrated to generate and output the corresponding verification guidance data.

2. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 1, characterized in that, A structured template and a signal standard knowledge base are constructed. The signal standard knowledge base is then vectorized to obtain a signal parameter semantic vector library, specifically as follows: The standardized form template's field layout, attribute types, and basic format requirements are analyzed to generate the structured template. Extract the vehicle-side signal parameters from the CAN bus communication protocol template and construct the signal standard knowledge base associated with the intelligent structured template fields; By calling the open-source semantic coding model interface, various parameters in the signal standard knowledge base are vectorized to generate the signal parameter semantic vector library.

3. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 1, characterized in that, The structured template includes at least one of the following: field attributes, data type, format specifications, field relationships, and validation priority.

4. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 2, characterized in that, The intelligent verification includes at least one of the following: basic format compliance verification, parameter consistency verification, mandatory item integrity verification, vehicle-side signal acquisition feasibility pre-verification, and cloud business rule compliance verification.

5. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 4, characterized in that, The parameter consistency check specifically involves: The signal parameters filled in by the intelligent structured template are vectorized and encoded through the open-source semantic coding model interface to obtain a temporary semantic vector. The temporary semantic vector is then matched with the standard parameter semantic vector in the signal parameter semantic vector library for similarity. If the similarity is within a preset threshold range, suggestions for modifying the filled content are provided.

6. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 4, characterized in that, The feasibility pre-verification of vehicle-side signal acquisition and the compliance verification of cloud business rules are both executed collaboratively through preset rule prompts and open-source large model interfaces, and the two verifications are completed synchronously through a single interface call.

7. The structured template pre-verification method for vehicle-side signal cloud uploading requirements according to claim 1, characterized in that, Also includes: The verification guidance data is displayed, which includes the marked error information and the guidance content on the basis for correction. After the marked error information is corrected, the verification is retried until all corresponding verification dimensions pass.

8. A structured template pre-verification device for vehicle-side signal cloud uploading requirements, characterized in that, include: The knowledge base and vector library construction module (1) is used to construct a structured template and a signal standard knowledge base based on the standardized form template for cloud-based vehicle-end signals of the car manufacturer and the CAN bus communication protocol template, and to perform vectorization encoding on the signal standard knowledge base to obtain a signal parameter semantic vector library. The intelligent template integration module (2) is used to associate and bind the structured template and the signal parameter semantic vector library to form an intelligent structured template, and to adapt and integrate the intelligent structured template with the corresponding form interaction plugin to realize the standardized filling interaction of the vehicle signal cloud access requirements; The intelligent verification execution module (3) is used to trigger the verification process based on the completion status of the intelligent structured template, and to perform intelligent verification on the user-filled content in the intelligent structured template in combination with the structured template and the signal parameter semantic vector library. The verification result processing module (4) is used to integrate and classify the various verification results output by the intelligent verification execution module (3), and generate and output the corresponding verification guidance data.

9. The structured template pre-verification device for vehicle-side signal cloud uploading requirements according to claim 8, characterized in that, The knowledge base and vector library construction module (1) includes: The structure graph construction unit (11) is used to parse the field layout, attribute types and basic format requirements of the standardized form template and generate the structured template. The knowledge base construction unit (12) is used to extract the vehicle-side signal parameters in the CAN bus communication protocol template and construct the signal standard knowledge base associated with the intelligent structured template fields; The semantic vector library generation unit (13) is used to call the open-source semantic coding model interface to perform vectorization transformation on various parameters in the signal standard knowledge base and generate the signal parameter semantic vector library.

10. The structured template pre-verification device for vehicle-side signal cloud uploading requirements according to claim 8, characterized in that, It also includes a display and loop verification trigger module (5), which is used to display the verification guidance data, which includes the guidance content of marked error information and correction basis, and re-triggers the verification after the marked error information is corrected, until all corresponding verification dimensions pass.

Images