Food research and development compliance auditing method and system based on category rigidity rule engine

Abstract

This invention discloses a food R&D compliance audit method, system, and integrated device based on a category-based rigid rule engine. The method achieves an automated closed loop from requirement input to compliant output in food R&D through structured R&D requirement input, four-layer hierarchical mandatory verification, chained R&D task orchestration, multi-gradient compliant formula generation, process parameter matching, intellectual property risk avoidance, filing document generation, and encrypted delivery and evidence storage. This invention can reduce non-compliant formula output and improve the consistency between formulas, processes, and label documents under the constraints of food regulations and national standards.

Description

Technical Field

[0001] This invention relates to the field of food R&D compliance auditing technology, specifically to a food R&D compliance auditing method and system based on a category-based rigid rule engine. Background Technology

[0002] In the food industry, new product development is a highly complex, long-term, and extremely risky systematic project. Researchers not only need to focus on the product's taste, flavor, and nutritional value, but also must strictly adhere to a series of mandatory national standards regarding food additives, nutritional labeling, and production licenses (SC). Currently, small and medium-sized food enterprises generally face the following difficulties when developing new products: First, there are numerous regulations and standards that are frequently updated. Researchers need to manually consult a large number of documents such as GB 2760 (Standard for the Use of Food Additives) and GB 14880 (Standard for the Use of Nutritional Fortifiers), which is not only inefficient but also highly susceptible to human error leading to formula violations, resulting in the risk of failure to register the product or its removal from the market.

[0003] Secondly, there is a disconnect between formulation design and process. In the traditional R&D model, the formulations designed by formulators often lack consideration for the compatibility with production equipment, leading to frequent problems such as mismatched process parameters and poor product stability in the pilot stage. Repeated debugging greatly increases R&D costs and time.

[0004] Secondly, intellectual property risk assessment is often delayed. Most companies only conduct patent searches towards the end of product development. Once infringement risks are discovered, the original design often needs to be scrapped, resulting in huge sunk costs.

[0005] Finally, existing auxiliary tools are mostly single query software or isolated hardware devices, lacking a fully automated solution covering the entire process from formula input, compliance review, process generation to label printing. Therefore, there is an urgent need in the market for an intelligent method and system that can deeply integrate regulatory logic, formula algorithms, and production control. Summary of the Invention

[0006] The present invention aims to solve the above-mentioned technical problems by providing a food R&D compliance audit method and system based on a category-based rigid rule engine. The present invention aims to provide a food R&D compliance audit method, system, and integrated device based on a category-based rigid rule engine to solve the aforementioned technical problems.

[0007] On the one hand, this invention provides a food R&D compliance review method based on a category-based rigid rule engine, including the following steps: S1, receive structured new product development requirements parameters entered by the user. The structured new product development requirements parameters include at least the food category, target population, raw material range, nutritional or flavor objectives, process equipment conditions, and label claims.

[0008] S2 calls the first-layer requirement compliance pre-inspection and interception layer to perform verification on the structured new product R&D requirement parameters for raw material legality, additive compliance, population suitability and claim compliance, and outputs a preliminary compliance risk screening report.

[0009] S3 unlocks the research and development tasks arranged in a fixed chain after the user's permission verification is passed.

[0010] S4 calls the second-layer formula component limit balance calculation layer to run a multi-objective balance algorithm under the national food safety standard threshold, enterprise standard threshold and raw material property constraints to generate at least one compliant formula that is lower than the national standard upper limit.

[0011] S5, invoke the third-layer process parameter rigid matching constraint layer, and generate the mass production process SOP based on the raw material properties and production line equipment parameters of the compliant formula.

[0012] S6 invokes the fourth layer of intellectual property risk mandatory avoidance layer, searches the formula and process patent database, and adjusts the ratio, raw material substitution parameters or process route to generate differentiated solutions when identifying infringement risks.

[0013] S7 generates enterprise standard texts, SC filing application materials, and compliance label documents based on formulas and processes that have undergone four-level progressive verification.

[0014] S8 encrypts and packages the formula, process SOP, filing documents, label text, rule hit logs and permission logs for delivery, and generates traceable operation logs and data asset ownership information.

[0015] Furthermore, in the four-layer hierarchical mandatory verification, the output of the previous layer serves as the input constraint for the next layer. If any layer outputs a non-compliant conclusion, subsequent formula or process generation is halted. Its advantage lies in establishing a strict hierarchical verification chain, preventing the downward transmission of non-compliant data, and eliminating the risk of violations at the source.

[0016] Furthermore, the multi-objective balancing algorithm uses cost, nutritional indicators, sensory objectives, additive limits, and process adaptability as objectives or constraints to output at least one of the following: a cost-effective mass-production formula, a highly nutritious and compliant formula, or a high-end efficacy formula. Its advantage lies in its ability to provide users with multiple compliant options at once, based on different market positioning and cost budgets, thus shortening the decision-making cycle in the product planning stage.

[0017] Furthermore, the mass production process SOP includes the feeding sequence, heating curve, holding time, homogenization speed, cooling gradient, sterilization parameters, and CIP cleaning parameters. Its advantage lies in the deep integration of process parameters with the formulation properties, avoiding multiple pilot-scale reworks caused by traditional experience-based process debugging, and achieving seamless integration from formulation to production.

[0018] On the other hand, the present invention provides a food R&D compliance audit system based on a category rigid rule engine, including a structured demand input module, a legal and national standard knowledge base module, a category rigid rule engine module, a chain-like skill execution module, a permission and work order scheduling module, a formula component limit balancing module, a process parameter matching module, an intellectual property risk avoidance module, a filing document generation module, and an encrypted delivery evidence storage module; each module is configured to execute the method described in any of the above.

[0019] Furthermore, the category-specific rigid rule engine module includes a demand compliance pre-inspection and interception layer, a formula component limit balance calculation layer, a process parameter rigid matching constraint layer, and an intellectual property risk mandatory avoidance layer. Its advantage lies in the fact that the four engine layers are progressive and mutually dependent, forming a rigorous compliance review chain that ensures the final output formula and process are flawless in terms of regulations, technology, and intellectual property.

[0020] On the other hand, the present invention provides a food R&D compliance audit integrated device based on a category rigid rule engine, including a cabinet, a human-machine interface, an industrial controller, a label output module, a data communication interface, and an authorization module; the industrial controller has a built-in memory and a processor, the memory stores a food regulation database and national standard rigid threshold data, and the processor is configured to execute the method described in any of the above.

[0021] Furthermore, the data communication interface is at least one of an RS485 interface, a CAN bus interface, or an Ethernet interface, used to send formula data and process parameters to external production equipment or a MES system; the label output module is used to print approved compliant label information. Its advantages lie in the fact that multiple communication interfaces cover mainstream industrial fieldbus protocols, ensuring strong compatibility, allowing for immediate label approval and printing, and eliminating compliance risks caused by label errors.

[0022] On the other hand, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any of the above-mentioned embodiments.

[0023] Compared with the prior art, the present invention has the following significant advantages: First, the entire process is automated and standardized. This invention integrates the previously fragmented processes of formula design, compliance review, process development, patent search, and label printing into a closed-loop process through the collaborative design of the method, system, and integrated device. Users only need to input basic requirements, and the system can automatically output compliant formulas, process specifications, and filing documents, greatly shortening the R&D cycle and reducing reliance on manpower.

[0024] Secondly, the system boasts extremely high regulatory compliance. Through its built-in rigid rule engine and national standard database, the system can perform real-time, item-by-item compliance checks on formulas. If a formula exceeds national standard limits or violates incompatibilities, the system will immediately intercept it and display an error message, fundamentally eliminating the risk of violations due to human negligence.

[0025] Third, it achieves seamless integration between R&D and production. The process standard operating procedure (SOP) generated by this invention not only includes theoretical parameters but can also be adapted and adjusted according to the characteristics of specific production equipment. It can also be directly sent to the production line via a data communication interface, avoiding repeated debugging in the pilot stage and significantly reducing R&D costs.

[0026] Fourth, it possesses commercial and legal protection functions. The authorization and authentication module provides the technical foundation for the commercial operation of the software service; while the built-in patent search and circumvention design functions can effectively reduce the risk of intellectual property infringement for enterprises and protect their innovative achievements. Attached Figure Description

[0027] Figure 1 The system flowchart provided for the embodiments of the present invention.

[0028] Figure 2 The system control diagram provided for an embodiment of the present invention. Detailed Implementation

[0029] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0030] Example 1: Audit Method and Flow like Figure 1 As shown in the figure, this embodiment provides a food R&D compliance audit method based on a category rigid rule engine.

[0031] In this embodiment, specifically: Step S1, the system receives structured new product development requirement parameters entered by the user. The user fills out a standardized work order through a human-computer interaction interface. The work order fields include at least the food category (e.g., modified milk powder, solid beverage), target population (e.g., infants, adults), raw material range (e.g., core raw materials, prohibited raw materials), nutritional or flavor objectives (e.g., high calcium, low sugar), process equipment conditions (e.g., existing production line model, capacity limitations), and label claims (e.g., "no additives" "rich in dietary fiber"). These parameters constitute the basic input for all subsequent verification and calculations.

[0032] In this embodiment, specifically in step S2, the system calls the first-layer requirement compliance pre-inspection and interception layer to perform four checks on the structured new product R&D requirement parameters. The checks include: raw material legality (checking whether prohibited substances outside the list of food and medicine homology are used); additive compliance (checking whether the additive combination complies with the inclusion principles and scope of use of GB 2760); population suitability (checking whether the formula ingredients are suitable for the target population (e.g., certain sweeteners are prohibited in infant food); and claim compliance (checking whether the label claims have sufficient scientific basis or do not exceed the permitted scope of regulations). After the checks are completed, a preliminary compliance risk screening report is output, with green, yellow, and red indicating passed, warning, and non-compliant items, respectively.

[0033] In this embodiment, specifically: Step S3, the system detects the user's permission verification status. When the user has not completed payment or the authorization has expired, the system only displays the preliminary screening report and locks the subsequent steps; when the user passes the permission authentication module verification, the system automatically unlocks the R&D tasks arranged in a fixed order, including formula generation, process preparation, patent search, and document output.

[0034] In this embodiment, specifically in step S4, the system calls the second-layer formula component limit balancing calculation layer. This layer runs a multi-objective balancing algorithm under the constraints of national food safety standard thresholds, enterprise standard thresholds, and raw material properties. The algorithm takes minimizing formula cost, maximizing nutritional indicators, and optimizing sensory taste as its objective functions, while strictly limiting the mass fraction of all additive components to below the maximum allowable limit value of the national standard. The system outputs at least one compliant formula at a time, preferably providing three tiered options for the user to choose from: a high-cost-performance mass production formula, a high-nutrition compliant formula, and a high-end efficacy formula.

[0035] In this embodiment, specifically: Step S5, the system calls the third-layer process parameter rigid matching constraint layer. Based on the raw material properties (such as viscosity, emulsifiability, and heat sensitivity) of the formula selected in step S4, suitable production line equipment is automatically matched from the built-in process equipment library, and a detailed mass production process SOP is generated. The SOP includes the feeding sequence, heating curve, holding time, homogenization speed, cooling gradient, sterilization parameters, and CIP cleaning parameters to ensure that the process parameters accurately correspond to the formula characteristics.

[0036] In this embodiment, specifically in step S6, the system invokes the fourth layer of mandatory intellectual property risk avoidance. The system automatically searches the built-in formula and process patent database and compares the current solution with the scope of claims of existing patents. When a potential infringement risk is identified, the system automatically adjusts the proportions of non-critical raw materials, replaces equivalent raw materials, or adjusts the process route without changing the core quality of the product, generating a differentiated solution without infringement, and prompts the user with avoidance details on the interface.

[0037] In this embodiment, specifically in step S7, based on the final formula and process that have passed four levels of progressive verification, the system automatically generates enterprise standard text, SC filing application materials, and compliant label documents that conform to the format required by the State Administration for Market Regulation. The label documents include all legally required information such as ingredient list, nutrition facts, production date, shelf life, storage conditions, and the SC mark.

[0038] In this embodiment, specifically in step S8, the system encrypts and packages the formula, process SOP, filing documents, label text, rule hit log, and permission log to generate a computation log containing a timestamp and operator ID, and completes the registration of data asset ownership through blockchain technology. The encrypted package can be transmitted to the enterprise server via the network or directly distributed to the production equipment through the data communication interface.

[0039] Example 2: System Module Structure like Figure 2 As shown in the figure, this embodiment provides a food R&D compliance review system based on a category rigid rule engine.

[0040] In this embodiment, specifically: the system includes a structured requirements input module, used to receive structured new product development requirements parameters entered by the user and format them into a standard data structure for transmission to subsequent modules. The regulations and national standards knowledge base module stores a complete food regulations database, including structured data from core regulations such as GB 2760, GB 14880, and GB 7718, as well as a raw material property database and a process equipment parameter database covering 12 major categories.

[0041] In this embodiment, specifically: the category rigid rule engine module is the core of this system, comprising four sub-modules. The demand compliance pre-inspection and interception layer is responsible for performing preliminary verifications such as raw material legality and additive compliance; the formula component limit balance calculation layer runs a multi-objective balance algorithm to generate compliant formulas under national standard threshold constraints; the process parameter rigid matching constraint layer automatically matches process parameters based on the formula's physical properties; and the intellectual property risk mandatory avoidance layer searches the patent database and generates differentiated solutions. A strict input-output dependency relationship exists between the four layers; failure of verification in any layer will terminate the subsequent process.

[0042] In this embodiment, specifically: the chained skill execution module schedules the execution of various R&D tasks according to a preset fixed order, including new product requirement compliance pre-inspection, multi-gradient compliance formula generation, pilot production process SOP preparation, cost and gross profit calculation, patent infringement risk investigation and avoidance, and automatic generation of enterprise standard / SC filing documents. The permission and work order scheduling module is connected to the chained skill execution module and controls the locking and unlocking of each task according to the user's permission status.

[0043] In this embodiment, specifically, the formulation component limit balancing module, process parameter matching module, intellectual property risk avoidance module, and filing document generation module correspond to the functions of each layer of the category rigid rule engine, and are the specific execution units of each layer's algorithm. The encrypted delivery and evidence storage module is responsible for encrypting and packaging the final results, generating operation logs, and completing the data asset ownership confirmation.

[0044] Example 3: Structure of the all-in-one device This embodiment provides an integrated device for food R&D compliance review based on a category-based rigid rule engine.

[0045] In this embodiment, specifically: the cabinet is a vertical cabinet made of stainless steel, with the interior divided into upper and lower installation spaces. The upper space is used to install the human-machine interface and tag output module, while the lower space is used to install the industrial controller and power module. Four casters are installed at the bottom of the cabinet for easy movement and positioning.

[0046] In this embodiment, specifically: the human-machine interface uses a 15.6-inch industrial-grade touch screen, which is hinged to the side of the cabinet and can be adjusted from 0° to 120°. The touch screen is used for users to input food formula parameters and display audit results, including compliance audit reports, formula cost analysis, and previews of enterprise standard / SC filing documents.

[0047] In this embodiment, specifically: the industrial controller is installed inside the cabinet, using an embedded industrial motherboard, equipped with a high-performance processor and a large-capacity solid-state drive. The memory is pre-loaded with a complete food regulations database and national standard rigid threshold data. The processor is configured to execute the auditing method described in Embodiment 1, receiving formula parameters from the human-machine interface, calling the category rigid rule engine to perform four-layer progressive mandatory verification, and generating a compliant formula after the audit is passed.

[0048] In this embodiment, specifically: the label output module uses an industrial-grade thermal transfer or thermal label printer, installed on the side or front panel of the cabinet, for printing approved compliant label information. The data communication interface is at least one of an RS485 interface, a CAN bus interface, or an Ethernet interface, installed on the back panel of the cabinet, for transmitting formula data and process parameters to external production equipment or the MES system.

[0049] In this embodiment, specifically: the authorization module is electrically connected to the industrial controller and is implemented using a hardware encryption chip, storing the user's license information and authorization period. When the user has not completed payment, the authorization module locks the industrial controller's recipe generation and process planning functions, only enabling the demand pre-inspection function.

[0050] Comparative Example 1 This comparative study employs a traditional manual review method for food R&D compliance audits. R&D personnel first consult paper or PDF versions of regulations and standards such as GB 2760, manually recording the maximum usage limits for each additive. Then, they manually calculate in an Excel spreadsheet whether the content of each component exceeds the limits. Next, they develop production process parameters based on experience, gradually adjusting them through multiple pilot-scale experiments. Finally, they commission external agencies to conduct patent searches and prepare filing documents. The entire process requires collaboration among multiple roles, including R&D engineers, regulatory specialists, process engineers, and intellectual property consultants, taking an average of about four weeks and is prone to human error.

[0051] Comparative Example 2 This comparative study utilizes existing single-function compliance query software for auxiliary review. This software only provides electronic query functions for regulations and standards; users need to determine the compliance of the formula themselves, as the software lacks automatic calculation and verification capabilities. Process parameters still need to be manually defined, and patent searches and filing document preparation rely on external professional institutions. Although the entire process is slightly faster than a purely manual method, it still takes approximately two weeks, and there is a significant amount of manual handover and repetitive confirmation work between each stage.

[0052] Examples 1 to 3 of this invention are significantly superior to Comparative Examples 1 and 2 in terms of total time consumption, number of formulation schemes, efficiency of process SOP generation, patent search coverage, ability to generate filing documents, data asset ownership confirmation, human resource input, and compliance omission rate. Although Example 3 takes slightly longer than Examples 1 and 2, it provides a complete hardware integration solution, achieving full-chain coverage from software algorithms to physical devices.

[0053] In summary, this invention, through the collaborative design of the method, system, and integrated device, combined with the step-by-step verification of the four-layer rigid rule engine and the chain-like task orchestration, achieves automation and standardization of the entire food R&D process, significantly shortens the R&D cycle, reduces compliance risks, and provides strong technical support for enterprise digital transformation.

[0054] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A food R&D compliance review method based on a category-based rigid rule engine, characterized in that, Includes the following steps: S1, receive structured new product development requirements parameters entered by the user. The structured new product development requirements parameters include at least the food category, target population, raw material range, nutritional or flavor objectives, process equipment conditions and label claims. S2, invoke the first layer of requirement compliance pre-inspection and interception layer to perform verification of the legality of raw materials, compliance of additives, suitability for the target population and compliance of claims on the structured new product R&D requirement parameters, and output a preliminary compliance risk screening report; S3 unlocks the R&D tasks arranged in a fixed chain after the user's permission verification is passed; S4 calls the second-layer formula component limit balance calculation layer to run a multi-objective balance algorithm under the national food safety standard threshold, enterprise standard threshold and raw material property constraints to generate at least one compliant formula below the national standard upper limit; S5, invoke the third-layer process parameter rigid matching constraint layer to generate mass production process SOP based on the raw material properties and production line equipment parameters of the compliant formula; S6 invokes the fourth layer of intellectual property risk mandatory avoidance layer, searches the formula and process patent database, and adjusts the ratio, raw material substitution parameters or process route to generate differentiated solutions when identifying infringement risks. S7 generates enterprise standard text, SC filing application materials, and compliance label text based on formulas and processes that have undergone four-level progressive verification. S8 encrypts and packages the formula, process SOP, filing documents, label text, rule hit logs and permission logs for delivery, and generates traceable operation logs and data asset ownership information.

2. The method according to claim 1, characterized in that, In the four-layer progressive mandatory verification, the output of the previous layer serves as the input constraint for the next layer. If any layer outputs an invalid conclusion, the subsequent formula generation or process generation is terminated.

3. The method according to claim 1, characterized in that, The multi-objective balancing algorithm takes cost, nutritional indicators, sensory objectives, additive limits, and process adaptability as objectives or constraints, and outputs at least one of the following: cost-effective mass production formula, high nutritional compliance formula, or high-end efficacy formula.

4. The method according to claim 1, characterized in that, The mass production process standard operating procedure (SOP) includes the feeding sequence, heating curve, holding time, homogenization speed, cooling gradient, sterilization parameters, and CIP cleaning parameters.

5. A food R&D compliance review system based on a category-based rigid rule engine, characterized in that, It includes a structured requirements input module, a regulations and national standards knowledge base module, a category rigid rules engine module, a chain-like skill execution module, a permission and work order scheduling module, a formula component limit balance module, a process parameter matching module, an intellectual property risk avoidance module, a filing document generation module, and an encrypted delivery evidence storage module; each module is configured to perform the method described in any one of claims 1 to 4.

6. The system according to claim 5, characterized in that, The category rigid rule engine module includes a demand compliance pre-inspection and interception layer, a formula component limit balance calculation layer, a process parameter rigid matching constraint layer, and an intellectual property risk mandatory avoidance layer.

7. A food R&D compliance review integrated machine device based on a category-based rigid rule engine, characterized in that, It includes a cabinet, a human-machine interface, an industrial controller, a tag output module, a data communication interface, and an authorization module; the industrial controller has a built-in memory and a processor, the memory stores a food regulations database and national standard rigid threshold data, and the processor is configured to execute the method described in any one of claims 1 to 4.

8. The all-in-one device according to claim 7, characterized in that, The data communication interface is at least one of RS485 interface, CAN bus interface or Ethernet interface, used to send formula data and process parameters to external production equipment or MES system; the label output module is used to print approved compliant label information.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 4.

Images