A method, apparatus, and medium for fragrance inspection monitoring

Abstract

The present application relates to the technical field of quality supervision of essence and flavor production. Specifically relates to a kind of for essence and flavor inspection monitoring method, device and medium, its method includes to target raw material information executes preprocessing operation, obtains preprocessed raw material information, according to formula requirement to preprocessed raw material information is weighed operation, obtains weighed raw material group, to weighed raw material group executes preparation operation, obtains preparation raw material combination information, to preparation raw material combination information executes mixing processing operation, obtains mixed product, to mixed product executes quality detection, obtains quality detection result, if quality detection result meets the quality index requirement set, then call the packaging material information corresponding to preparation raw material combination information, to mixed product executes packaging processing operation, obtains finished product information.The present application has the effect of solving the quality supervision ability shortage in essence and flavor preparation process.

Description

Technical Field

[0001] This invention relates to the technical field of quality supervision in the production of fragrances and flavorings, and in particular to a method, apparatus and medium for testing and monitoring fragrances and flavorings. Background Technology

[0002] In traditional fragrance and flavor production, raw material management typically relies on paper records or simple spreadsheets to document basic information, lacking effective traceability. Raw material pretreatment generally involves experience-based cleaning and crushing. Weighing and mixing are largely done manually according to the formula, with manual reading of measuring instruments and manual addition and mixing of materials. Final product quality inspection is often done through sampling, with limited testing items, making it difficult to comprehensively and accurately assess product quality. Packaging is typically done according to common packaging specifications.

[0003] However, existing preparation methods have significant drawbacks. Traditional raw material management methods cannot effectively trace the source of raw materials, making it difficult to pinpoint the root cause when product quality issues arise. Manual weighing and preparation are prone to errors, leading to inconsistent product quality. Sampling inspections cannot guarantee that every batch meets quality requirements, potentially allowing substandard products to enter the market. Furthermore, the lack of traceability and feedback mechanisms corresponding to raw materials and processing procedures makes it impossible to promptly identify and resolve potential problems in the production process. Summary of the Invention

[0004] To address the problem of insufficient quality supervision during the preparation of fragrances and flavors, this application provides a method, apparatus, and medium for testing and monitoring fragrances and flavors.

[0005] The above-mentioned objective of this application is achieved through the following technical solution:

[0006] A method for testing and monitoring flavorings and fragrances, the method comprising:

[0007] Obtain the raw material information required for the preparation of fragrances and flavors, conduct incoming inspection of the raw material information, generate a traceability identification code after the raw material information passes inspection, and establish a binding relationship between the traceability identification code and the raw material information to obtain the target raw material information;

[0008] Perform a preprocessing operation on the target raw material information to obtain preprocessed raw material information;

[0009] Weigh the pretreated raw materials according to the formula requirements to obtain a weighed raw material group. Perform a preparation operation on the weighed raw material group to obtain the preparation raw material combination information.

[0010] Perform a mixing process on the raw material combination information to obtain the mixed product;

[0011] The product after the mixing process is subjected to quality testing to obtain the quality testing results;

[0012] If the quality test results meet the set quality index requirements, then the packaging material information corresponding to the raw material combination information is called, and the product after mixing is packaged to obtain the finished product information.

[0013] By adopting the above technical solution, and generating traceability identification codes and establishing a binding relationship with raw material information after the raw materials have passed inspection, digital traceability of the source of raw materials can be achieved. This effectively solves the problem that traditional paper-based record-keeping methods cannot accurately trace the source. By performing preprocessing, weighing, and preparation operations on the target raw material information, the raw material processing and feeding process is standardized, reducing human error and improving preparation accuracy. By performing mixing and quality inspection, it is ensured that each batch of products undergoes full quality inspection, avoiding the problem of insufficient quality coverage caused by traditional sampling inspection mechanisms. If the quality inspection is qualified, the packaging material information corresponding to the raw material combination information is further called for packaging, realizing automatic matching of packaging selection and process tracking. Overall, by linking raw materials, formulas, process parameters, test results, and packaging operations into one, traceability, quantification, and controllability from raw material input to finished product output are achieved.

[0014] In a preferred embodiment, this application can be further configured as follows: establishing a binding relationship between the traceability identifier and the raw material information to obtain the target raw material information includes:

[0015] Perform a field-level append operation on the traceability identifier and the raw material information, and add a field position in the raw material information to store the traceability identifier;

[0016] Write the traceability identifier into the field location and perform binding status registration. The raw material information that completes the binding relationship is the target raw material information.

[0017] By adopting the above technical solution, traceability identification codes can be accurately embedded in the raw material information data structure at the field level. By adding a new field position, the raw material information and traceability identification codes are statically bound together. Combined with the binding status registration mechanism, the consistency record of raw material identification and status is completed, thereby ensuring that each batch of raw materials has a unique traceable identification, so that the source and usage path of raw materials can be accurately located in the subsequent production process.

[0018] In a preferred embodiment, this application may be further configured such that: performing quality testing on the product after the mixing process to obtain a quality testing result further includes:

[0019] The quality test is performed on the product after the mixing process to obtain mixing process parameter information;

[0020] The mixing process parameter information is input into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing operation. If there is a process deviation in the mixing operation, an abnormal prompt message is generated.

[0021] By adopting the above technical solution, key parameter information of the mixing process can be obtained synchronously after the mixing operation is performed, and this information can be input into the trained processing behavior monitoring model for intelligent analysis. The model has the ability to identify abnormal process modes by learning from historical behavior data, so it can determine in real time whether there is a behavior deviation in the current mixing operation. Once an operation behavior that deviates from the normal process range is detected, abnormal prompt information can be generated in a timely manner, realizing full-process monitoring of the mixing process quality.

[0022] In a preferred embodiment, this application can be further configured such that, before inputting the mixed process parameter information into the trained processing behavior monitoring model, it also includes:

[0023] Obtain historical mixing process parameter information, and construct a multidimensional training sample set containing behavioral bias labels based on the historical mixing process parameter information;

[0024] The multidimensional training sample set is input into the processing behavior monitoring model to be trained to extract time-series distribution features and fluctuation range features;

[0025] Based on the time-series distribution characteristics and the fluctuation interval characteristics, a feature discrimination function is constructed to characterize the boundary of behavioral stability.

[0026] Based on the feature discrimination function, the behavioral deviation types of each sample in the multidimensional training sample set are classified and labeled to generate training supervision information.

[0027] Based on the training supervision information, a supervised learning algorithm is used to update the parameters of the processing behavior monitoring model to be trained, thereby obtaining the trained processing behavior monitoring model.

[0028] By adopting the above technical solution, a multi-dimensional training sample set with behavioral deviation annotations can be constructed based on historical mixed process parameter information. This allows the processing behavior monitoring model to be trained to not only learn the temporal change patterns and fluctuation characteristics of each parameter during the training phase, but also accurately delineate the behavioral stability boundary through the constructed feature discrimination function. This enables effective classification of different deviation types and generation of supervision information. Based on this, combined with the constructed training supervision information, the model is gradually optimized and its structure adjusted using a supervised learning algorithm. Ultimately, a processing behavior monitoring model with high-precision process deviation identification capability is obtained, solving the problems of lack of specificity in model training, weak generalization ability, and inability to cope with behavioral diversity in existing technologies.

[0029] In a preferred embodiment, this application can be further configured as follows: the mixing process parameter information is input into a trained processing behavior monitoring model to determine whether there is a process deviation in the mixing operation; if there is a process deviation in the mixing operation, an abnormal prompt message is generated, including:

[0030] Extract the time-varying trends of each parameter in the mixing process parameter information, and construct a set of trend features;

[0031] The trend feature set is matched with the preset process behavior patterns in the trained processing behavior monitoring model to obtain the matching results;

[0032] Identify whether there are parameter combinations in the matching results that continuously deviate from the process behavior pattern; if so, generate the abnormal prompt information.

[0033] By adopting the above technical solution, a complete set of trend features can be constructed based on the changing trends of each parameter in the mixing process parameter information over time, thereby fully expressing the dynamic evolution path of each process parameter in the mixing process. Then, by matching this set of trend features with the process behavior patterns set in the pre-trained processing behavior monitoring model one by one, the matching deviation and behavior consistency degree are calculated. On this basis, parameter combinations that have continuous deviations in the matching process are identified, and abnormal prompt information is generated accordingly, thereby realizing early perception and accurate positioning of abnormal process behavior.

[0034] In a preferred embodiment, this application can be further configured as follows: if the quality test result meets the set quality index requirements, then the packaging material information corresponding to the raw material combination information is called to perform a packaging process on the mixed product to obtain finished product information, and the application further includes:

[0035] If the quality test results do not meet the quality index requirements, the target raw material information and processing node information associated with the mixed product are determined based on the traceability identification code to obtain abnormal record information;

[0036] Based on the anomaly record information, a corresponding feedback processing instruction is generated.

[0037] By adopting the above technical solution, when the quality inspection results do not meet the set index requirements, the traceability identification code can be used to quickly locate the raw material information associated with the current non-conforming product and its specific processing trajectory in each processing node, forming anomaly record information containing the source of raw materials and process details. Based on this, corresponding feedback processing instructions are generated according to the anomaly record information, thereby realizing systematic analysis and targeted processing of the source of the anomaly.

[0038] In a preferred embodiment, this application can be further configured such that: generating corresponding feedback processing instructions based on the anomaly record information includes:

[0039] By using a causal source analysis mechanism, the target raw material information and the processing node information in the abnormal record information are extracted to obtain a set of dominant factors;

[0040] Based on the set of dominant factors, a matching optimization strategy is retrieved from the process rule base, and the feedback processing instruction is generated in combination with the current production status parameters.

[0041] By adopting the above technical solution, based on the target raw material information and processing node information in the abnormal record information, the key factors that have a dominant influence on the abnormal output results can be identified through the causal tracing analysis mechanism. Then, the optimization strategy that matches the set of dominant factors can be accurately retrieved in the process rule base, and adaptive adjustments can be made in combination with the current production status parameters to generate targeted feedback processing instructions, thereby achieving effective response and improvement to abnormal process behavior.

[0042] The second objective of this invention is achieved through the following technical solution:

[0043] A device for testing and monitoring flavorings and fragrances, the device comprising:

[0044] The raw material binding module is used to obtain the raw material information required for the preparation of flavorings and fragrances, to inspect the raw material information upon receipt, to generate a traceability identification code after the raw material information has passed inspection, and to establish a binding relationship between the traceability identification code and the raw material information to obtain the target raw material information.

[0045] The preprocessing module is used to perform preprocessing operations on the target raw material information to obtain preprocessed raw material information;

[0046] The preparation module is used to weigh the pretreated raw material information according to the formula requirements to obtain the weighed raw material group, and to perform a preparation operation on the weighed raw material group to obtain the preparation raw material combination information.

[0047] The mixing module is used to perform a mixing operation on the combination information of the raw materials to obtain the mixed product.

[0048] The quality inspection module is used to perform quality inspection on the product after the mixing process and obtain the quality inspection results;

[0049] The packaging execution module is used to, if the quality inspection results meet the set quality index requirements, call the packaging material information corresponding to the raw material combination information to perform packaging processing operations on the mixed product and obtain finished product information.

[0050] By adopting the above technical solution, and generating traceability identification codes and establishing a binding relationship with raw material information after the raw materials have passed inspection, digital traceability of the source of raw materials can be achieved. This effectively solves the problem that traditional paper-based record-keeping methods cannot accurately trace the source. By performing preprocessing, weighing, and preparation operations on the target raw material information, the raw material processing and feeding process is standardized, reducing human error and improving preparation accuracy. By performing mixing and quality inspection, it is ensured that each batch of products undergoes full quality inspection, avoiding the problem of insufficient quality coverage caused by traditional sampling inspection mechanisms. If the quality inspection is qualified, the packaging material information corresponding to the raw material combination information is further called for packaging, realizing automatic matching of packaging selection and process tracking. Overall, by linking raw materials, formulas, process parameters, test results, and packaging operations into one, traceability, quantification, and controllability from raw material input to finished product output are achieved.

[0051] The above-mentioned objective three of this application is achieved through the following technical solution:

[0052] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for testing and monitoring fragrances and flavors.

[0053] The fourth objective of this application is achieved through the following technical solution:

[0054] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for testing and monitoring fragrances and flavors.

[0055] In summary, this application includes at least one of the following beneficial technical effects:

[0056] 1. By generating traceability identification codes and establishing a binding relationship with raw material information after the raw materials have passed inspection, digital traceability of the source of raw materials can be achieved, effectively solving the problem that traditional paper record methods cannot accurately trace the source. By performing preprocessing, weighing and preparation operations on the target raw material information, the raw material processing and feeding process is standardized, reducing human operation errors and improving preparation accuracy. By performing mixing processing and quality inspection, it is ensured that each batch of products undergoes full quality inspection, avoiding the problem of insufficient quality coverage caused by traditional sampling inspection mechanisms. If the quality inspection is qualified, the packaging material information corresponding to the raw material combination information is further called for packaging, realizing automatic matching of packaging selection and process tracking. Overall, by linking raw materials, formulas, process parameters, test results and packaging operations into one, traceability, quantification and controllability from raw material input to finished product output are achieved.

[0057] 2. It can construct a complete set of trend features based on the changing trends of each parameter in the mixing process parameter information over time, thereby fully expressing the dynamic evolution path of each process parameter in the mixing process. Then, by matching this set of trend features with the process behavior patterns set in the pre-trained processing behavior monitoring model, the matching deviation and behavior consistency are calculated. Based on this, parameter combinations that have continuous deviations in the matching process are identified, and abnormal prompt information is generated accordingly, thereby achieving early perception and accurate positioning of abnormal process behavior.

[0058] 3. When the quality inspection results do not meet the set index requirements, the traceability identification code can be used to quickly locate the raw material information associated with the current non-conforming product and its specific processing trajectory in each process node, forming anomaly record information containing raw material source and process details. Based on this, corresponding feedback processing instructions are generated according to the anomaly record information, thereby realizing systematic analysis and targeted processing of the anomaly source. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of a method for testing and monitoring fragrances and flavors in one embodiment of the application;

[0060] Figure 2 This is a flowchart illustrating the implementation of step S10 in a method for testing and monitoring fragrances and flavors according to an embodiment of this application.

[0061] Figure 3 This is a flowchart illustrating the implementation of step S50 in a method for testing and monitoring fragrances and flavors according to an embodiment of this application.

[0062] Figure 4This is a flowchart illustrating the implementation of a method for testing and monitoring fragrances and flavors before step S502 in one embodiment of this application.

[0063] Figure 5 This is a flowchart illustrating the implementation of step S502 in a method for testing and monitoring fragrances and flavors according to an embodiment of this application.

[0064] Figure 6 This is a flowchart illustrating the implementation of step S60 in a method for testing and monitoring fragrances and flavors according to an embodiment of this application.

[0065] Figure 7 This is a flowchart illustrating the implementation of step S602 in a method for testing and monitoring fragrances and flavors according to an embodiment of this application.

[0066] Figure 8 This is a schematic diagram of a fragrance and flavor testing and monitoring device according to one embodiment of this application;

[0067] Figure 9 This is a schematic diagram of a device according to one embodiment of this application. Detailed Implementation

[0068] The present application will be further described in detail below with reference to the accompanying drawings.

[0069] In one embodiment, such as Figure 1 As shown, this application discloses a method for testing and monitoring fragrances and flavorings, which specifically includes the following steps:

[0070] S10: Obtain the raw material information required for the preparation of fragrances and flavors, conduct incoming inspection of the raw material information, generate a traceability identification code after the raw material information passes inspection, and establish a binding relationship between the traceability identification code and the raw material information to obtain the target raw material information.

[0071] Specifically, by reading the raw material number, name, batch number, supplier information, expiration date, and warehousing time recorded in the raw material warehousing list, and storing this information as structured raw material basic data, the system calls the acceptance rule template. Based on the raw material's physicochemical indicator test report and packaging status confirmation list, it checks each item against the requirements, including appearance, odor, moisture content, and purity. After completing the qualification judgment, a traceability identification code is constructed using a unique number generation algorithm combined with the timestamp, supply batch number, and acceptance personnel number. An extended field is added to the raw material basic data structure to accommodate this traceability identification code, and the code is written into this field. At the same time, a binding status identifier is registered in the raw material information management database, indicating that the raw material basic data has been logically associated with the traceability identification code. At this point, the raw material information with traceability capability after binding is the target raw material information.

[0072] S20: Perform a preprocessing operation on the target raw material information to obtain preprocessed raw material information.

[0073] In this embodiment, the pretreatment operation refers to the preliminary processing flow performed according to the type and physical properties of the raw materials. The processing flow includes washing, drying, screening or crushing the raw materials to remove impurities, reduce moisture content or adjust particle size distribution so that the raw materials meet the formulation and mixing requirements of subsequent preparation steps.

[0074] Specifically, based on the raw material type, physical form, and batch attributes contained in the target raw material information, the required pretreatment method for each raw material is determined. By searching for a preset processing rule set that matches the raw material type, the pretreatment steps and parameter configurations corresponding to each raw material are obtained. For example, for powdered raw materials, a drying operation is performed to control the moisture content within a set range, followed by sieving to remove particles with abnormal particle size. For liquid raw materials, homogenization or filtration is performed based on their density and viscosity information to ensure that their composition is evenly distributed and free of visible impurities. For viscous or resinous raw materials, the corresponding temperature control pretreatment parameters are read, and heating and dilution are performed to improve their fluidity. After each operation is completed, the information writing instruction is called to uniformly update the pretreatment status area of ​​the original fields in the raw material information with the post-processing parameter status, processing time, processing method identifier, and other information, thereby forming the pre-processed raw material information.

[0075] S30: Weigh the pretreated raw materials according to the formula requirements to obtain the weighed raw material group. Perform the preparation operation on the weighed raw material group to obtain the preparation raw material combination information.

[0076] In this embodiment, the weighing operation refers to the process of weighing the pretreated raw materials one by one according to the set mass value based on the component ratio requirements in the fragrance and flavor formula. Weighing can be completed by digital electronic scales, automatic weighing devices, or manual reading instruments. The obtained mass data of each group of raw materials is used to generate the corresponding weighed raw material group. The preparation operation refers to the process of combining the weighed raw materials according to the formula order and proportion requirements. The combination may include the pouring of raw materials, blending before stirring, and time-series controlled addition, etc., to generate the preparation raw material combination information for subsequent mixing.

[0077] Specifically, based on the raw material type, processing status, and batch identification contained in the pre-processed raw material information, the proportioning parameters of each component raw material are extracted according to the preset fragrance and flavor formula requirements, including the target addition amount and allowable error range. An electronic weighing device with a high-precision sensor is used to weigh each component individually. During the weighing process, the weight data of each raw material is collected in real time and compared with the corresponding proportioning parameters to determine if it is within the allowable error range. If the weighing result exceeds the range, removal or addition operations are performed until the formula requirements are met. After weighing, all raw materials are added to the mixing container in the prescribed order. Mechanical stirring is used during the preparation operation for low-speed pre-fusion to ensure uniform material addition and provide a consistent initial state for subsequent mixing operations. Stirring parameters include speed, duration, and interval. The pre-fusion process is controlled according to the operating conditions specified in the formula, and the material dispersion uniformity is judged by sampling at set times. After confirming that the preparation requirements are met, the preparation result is saved as the raw material combination information.

[0078] S40: Perform a mixing operation on the raw material combination information to obtain the mixed product.

[0079] In this embodiment, the mixing process refers to the uniform mixing of the raw material combination information by using a stirring device after the preparation is completed. The mixing process includes two parts: the main mixing stage and the final mixing stage. The main mixing stage refers to the process of performing high-intensity stirring on the pre-fused raw material combination information to achieve full dispersion. The final mixing stage refers to the process of maintaining low-speed stirring in the later stage of the mixing process to stabilize the material system and eliminate local concentration differences. By controlling the stirring speed, mixing time and temperature parameters, the mixing process ensures that the components in the raw material combination information are fully in contact in space and form a stable and uniform mixing state.

[0080] Specifically, based on the raw materials and their corresponding proportions, mixing order, and operating parameters recorded in the raw material combination information, the automated mixing equipment performs the mixing process. Following the prescribed order, the automated mixing equipment sequentially controls the feeding device to introduce the weighed raw material groups into the mixing chamber. The stirring device starts and enters the main mixing stage. The motor drives the stirring blades to rotate at a set speed. The set stirring speed value is extracted based on the raw material viscosity, target uniformity, and historical process parameters, and dynamically corrects the speed parameters based on the resistance data fed back by the sensors in real time. Dynamic speed parameter correction refers to the control process of continuously fine-tuning the stirring speed based on the real-time resistance change value collected by the sensors to maintain stable shear force. During the stirring process, intermittent time and reversal cycle are set to improve mixing uniformity. The temperature control device adjusts the internal temperature of the mixing chamber and maintains it within a suitable reaction range to avoid fluctuations in raw material properties. When the stirring time reaches the process set value and the torque change rate fed back by the sensors stabilizes within the preset threshold range, the stirring device enters the final mixing stage to maintain stable material distribution. After the mixing process is completed, the stirring device stops operating, the mixture is transferred to the output container, and the batch information is recorded. This mixture is the product after the mixing process.

[0081] S50: Perform quality testing on the product after mixing and obtain the quality test results.

[0082] In this embodiment, quality testing refers to the physicochemical property analysis and parameter measurement operations performed on the product after mixing. The test items include indicators such as viscosity, density, color, odor intensity or component ratio. The test methods can be a combination of instrument measurement, chemical analysis or manual olfaction evaluation.

[0083] Specifically, during quality testing, the corresponding testing items and parameter standard ranges are first retrieved based on the product category, formula type, and production batch number recorded in the raw material combination information. The testing process is then configured according to these standards. The mixed product sample is extracted manually or automatically, and component testing, physical parameter testing, and stability assessment are performed. Component testing uses gas chromatography to obtain the retention time and peak area data of the aroma active components, which are compared with the target concentration values ​​of each component in the standard formula. In physical parameter testing, an electronic densitometer and viscometer are used to measure density and flow characteristics, respectively. The difference between the measured values ​​and the allowable range of the formula is calculated using a built-in formula. In the stability assessment, the color and odor changes of the sample are observed within a specified time under a set environment. Significant change points are recorded, and deviations are judged against the set stability standards. After completing all testing operations, the collected raw data is normalized using statistical methods, and a comprehensive quality score is calculated according to weighted rules. The score is compared with the quality indicator requirements to determine whether the product meets the set quality standards, thus outputting the quality testing results.

[0084] S60: If the quality inspection result meets the set quality index requirements, call the packaging material information corresponding to the formulated raw material combination information, and perform a packaging operation on the product after the mixing process to obtain finished product information.

[0085] In this embodiment, the quality index requirements refer to the allowable range of various performance parameters set according to the flavor and fragrance product standards or user-customized specifications. The parameters include, but are not limited to, color uniformity, odor matching degree, component concentration, and stability, etc. The inspection result needs to fall entirely within the set range to be considered qualified. The packaging material information refers to a data set such as the container type, label style, sealing method, and packaging capacity corresponding to the formulated raw material combination information. The information is used to guide the selection of suitable packaging materials in the packaging operation and complete subsequent filling and labeling and other processing procedures.

[0086] Specifically, after confirming that the quality inspection result meets the set quality index requirements, according to the product number, usage type, and batch information included in the formulated raw material combination information corresponding to the product after the mixing process, retrieve the established packaging material database, determine the packaging specifications, container materials, sealing forms, and label templates matching this type of product, generate a packaging material call instruction based on the information, automatically complete the selection of packaging containers and material allocation, then send the qualified packaging containers to the packaging station, perform a filling operation, and the filling volume is set according to the product type and sales purpose. Control the single filling quality through an automatic filling device and weigh in real time to correct deviations. After filling, perform a sealing operation and attach the corresponding label. The label content is automatically generated according to the formulated raw material combination information, including product name, ingredient list, production batch number, shelf life, operator number, etc. After completing the above process, bind the unique identification code of each packaging unit with the product traceability information, and finally form finished product data with a complete information structure, which is recorded as finished product information.

[0087] Furthermore, the packaging material database refers to a structured data set used to store various packaging material information. The database records the packaging container types, packaging parameters, and usage rules suitable for flavor and fragrance products of different fragrances or uses, and supports rapid retrieval and call based on the formulated raw material combination information.

[0088] In one embodiment, as Figure 2 shown, in step S10, establish a binding relationship between the traceability identification code and the raw material information to obtain the target raw material information, including:

[0089] S101: Perform a field-level attachment operation on the traceability identification code and the raw material information, and add a field position for storing the traceability identification code in the raw material information.

[0090] Specifically, after completing the standardized entry of raw material information, based on the database field definition rules, a new field is added to the raw material information record using a structured data expansion method. This field is used to store the corresponding traceability identifier. When adding a new field, the field name, data type, length limit, and non-null constraint must be defined. The field name is set to an identifier consistent with the traceability logic, such as "trace_code". The data type is set to a variable-length character type to adapt to different encoding rules. The length is set to no less than 32 bits according to the traceability identifier structure to ensure complete encoding. After completing the field structure expansion, the generated traceability identifier is inserted into this field, and the record status is marked as "bound" in the raw material information data table so that the raw material record can be uniquely located in subsequent information flow and data consistency can be ensured.

[0091] S102: Write the traceability identifier into the field position and perform binding status registration. The raw material information that completes the binding relationship is the target raw material information.

[0092] Specifically, after adding the traceability identifier field, the generated traceability identifier is assigned to the preset field position in the raw material information record in a corresponding manner using a field write operation. During the write operation, the database interface function is called to perform an update operation on the specified field of the record to ensure that the traceability identifier matches the uniqueness of the raw material record. Then, the data status control module is called to attach a binding status identifier to the record. This identifier is set to "bound" in the status field to distinguish raw material data that has not completed traceability registration. The binding status is used to manage data integrity. After the field write and status update are completed, the raw material information containing the traceability identifier and the attached binding status is defined as the target raw material information.

[0093] In one embodiment, such as Figure 3 As shown, in step S50, which involves performing quality testing on the product after mixing to obtain the quality testing results, the method further includes:

[0094] S501: Perform quality testing on the product after mixing to obtain mixing process parameter information.

[0095] Specifically, during the quality inspection process, sensor devices are arranged both outside and inside the mixing container to synchronously collect key physical quantity data during the mixing process. These key physical quantity data include parameters such as stirring speed, stirring time, material temperature, mixing viscosity, conductivity, and power consumption. The signals collected by the sensors are converted into digital signals by an analog-to-digital converter and then input to the data acquisition device. The data acquisition device performs continuous sampling operations on each parameter at set time intervals to form parameter sequence data with timestamps. The data processing module performs normalization processing and outlier removal operations on the collected raw parameter data, and organizes and sorts the processed parameter data in a structured manner according to the mixing task number to generate mixing process parameter information with analyzable attributes.

[0096] S502: Input the mixing process parameter information into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing operation. If there is a process deviation in the mixing operation, generate an abnormal prompt message.

[0097] In this embodiment, the trained processing behavior monitoring model refers to a model trained based on labeled historical data of the mixed process, which is used to identify whether there is a behavioral deviation in the current mixed operation. The model can be a Long Short-Term Memory Neural Network (LSTM) that supports time-series modeling capabilities. Its input is a multi-dimensional mixed parameter sequence, and its output is the process stability status or anomaly identification result.

[0098] Specifically, the acquired mixed process parameter information is uniformly encoded into a feature vector sequence that conforms to the model input format. The parameter values ​​are arranged in chronological order and normalized by a feature mapping function before being input into a trained processing behavior monitoring model. The model is trained based on a supervised learning algorithm and has the ability to identify the differences between typical mixed behaviors and abnormal behaviors. After receiving the input, the model uses an internal multi-dimensional feature discrimination mechanism to match and score the input mixed parameter sequence with the process behavior template. Based on the set deviation threshold, it judges whether there is a deviation in the mixed processing operation that exceeds the normal process range. If the matching score is lower than the preset threshold, it is considered that there is a process deviation. Then, based on the identified deviation features, the abnormal identification generation module is triggered to extract key influencing factors and construct abnormal prompt information.

[0099] In one embodiment, such as Figure 4 As shown, before step S502, that is, before inputting the mixed process parameter information into the trained processing behavior monitoring model, the following steps are also included:

[0100] S5021: Obtain historical mixing process parameter information and construct a multidimensional training sample set containing behavioral bias labels based on the historical mixing process parameter information.

[0101] Specifically, mixing process parameter information for multiple completed batches of flavor and fragrance blends is collected. This information includes raw process variables such as stirring speed, temperature changes, viscosity changes, and raw material addition times recorded within the mixing time series. Based on these raw parameter data, the data is organized chronologically, missing values ​​are filled in, and outliers are removed or corrected. Then, through manual annotation or expert system analysis, the mixing status of each batch is labeled as "normal behavior" or "behavioral deviation exists," forming sample items with clear behavioral deviation labels. On this basis, feature combination technology is used to splice process parameters from different dimensions to construct a multidimensional training sample set that reflects the stability, abnormal trends, and dynamic response characteristics of mixing behavior.

[0102] S5022: Input the multidimensional training sample set into the processing behavior monitoring model to be trained, and extract the time-series distribution features and fluctuation range features.

[0103] Specifically, the constructed multidimensional training sample set is input into the processing behavior monitoring model to be trained. First, the sliding time window technique is used to segment each sample data to preserve the time dependency in the mixing process. Then, statistical analysis methods and convolution processing techniques are used to extract the time-series distribution characteristics of the process parameters, such as the mean, variance, peak value, and slope changes, within each time window. At the same time, by identifying the extreme points and inflection points of parameter changes, the range of numerical fluctuations in continuous intervals is analyzed to determine the fluctuation amplitude, duration, and frequency of change of each parameter under typical abnormal conditions. In this way, fluctuation interval characteristics reflecting different operational fluctuation states are extracted.

[0104] S5023: Based on time series distribution characteristics and fluctuation range characteristics, construct a feature discrimination function to characterize the boundary of behavioral stability.

[0105] Specifically, based on the extracted temporal distribution features and fluctuation range features, the behavioral feature vectors of all training samples are first projected into a unified standardized feature domain through multi-dimensional feature space mapping. Then, the support vector boundary construction method is used to fit the feature distribution regions labeled as normal behavior in the samples. The boundary judgment model is constructed by combining the distribution density of various features in the feature domain, the degree of edge mutation, and feature synergy. A regularization term is introduced to adjust the boundary sensitivity and generalization ability. On this basis, the behavioral stability boundary is defined as a classification hyperplane describing the deviation of the operation from the normal behavior threshold in the mixing process. This feature discrimination function can continuously evaluate the stability state of any unknown sample and calibrate the degree of deviation.

[0106] S5024: Based on the feature discrimination function, classify and label the behavioral deviation types of each sample in the multidimensional training sample set to generate training supervision information.

[0107] Specifically, based on the constructed feature discrimination function, each sample data in the multidimensional training sample set is traversed. First, the temporal distribution features and fluctuation range features of each sample are normalized and mapped to the behavioral stability boundary space defined by the feature discrimination function. Then, according to the judgment rules of the discrimination function, the position of the sample in the feature space is compared with the position of the boundary. If it falls inside the boundary, it is labeled as "stable behavior". If it falls outside the boundary, its deviation type is further judged based on its distance and direction beyond the boundary and its similarity with historical abnormal samples, and specific behavioral labels such as "insufficient mixing", "violent parameter fluctuation", and "abnormal rhythm" are assigned. Finally, the label results of all samples are combined with the corresponding input features to form the training supervision information required for supervised training.

[0108] S5025: Based on training supervision information, a supervised learning algorithm is used to update the parameters of the processing behavior monitoring model to be trained, thereby obtaining a trained processing behavior monitoring model.

[0109] Specifically, the Long Short-Term Memory (LSTM) network, which supports long-term dependency modeling, is selected as the training tool. The feature inputs from the training supervision information and the corresponding behavioral bias labels are used to form an input-output pair and fed into the processing behavior monitoring model to be trained. After initializing the model weight parameters, the model output is calculated based on the forward propagation mechanism, and the output is compared with the true labels. The prediction error is calculated using the cross-entropy loss function, and the gradient is calculated through the backpropagation algorithm. The gradient update operation is then performed in conjunction with the Adam optimizer to gradually optimize the weight parameters of each layer of the model. When the training error tends to stabilize on the validation set and meets the preset accuracy standard, the training process is completed and the trained processing behavior monitoring model is output.

[0110] In one embodiment, such as Figure 5 As shown, in step S502, the mixing process parameter information is input into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing operation. If there is a process deviation in the mixing operation, an abnormal prompt message is generated, including:

[0111] S5026: Extract the changing trends of each parameter over time from the parameter information of the mixing process, and construct a set of trend features.

[0112] Specifically, key parameter sequences, including stirring speed, stirring torque, material temperature, and mixing tank pressure, are extracted from the mixing process parameter information in time stamp order. For each type of parameter, the mean, maximum, minimum, and first derivative are calculated in each time period using a sliding window method. Furthermore, moving average and exponential weighted average methods are used to smooth each parameter sequence, capturing its short-term and long-term trends. Trend direction encoding is performed on each parameter sequence to numerically express trend patterns such as rising, falling, or oscillating. Finally, the trend features of all parameters are spliced ​​together to form a multi-dimensional trend feature set.

[0113] S5027: Match the set of trend features with the preset process behavior patterns in the trained processing behavior monitoring model to obtain the matching results.

[0114] Specifically, the constructed trend feature set is input into the process behavior pattern matching structure set in the trained processing behavior monitoring model. This structure is based on the temporal feature distribution of standard process behavior learned by the selected Long Short-Term Memory Network (LSTM) during the training phase. The input trend features are fed into the model step by step in chronological order. By comparing the similarity between the current trend feature and the standard process behavior path stored in the model's internal memory unit, the matching error value at each time step is calculated. The accumulated error score is used to determine whether the overall trend deviates from the standard process behavior trajectory. Finally, the output includes the score result containing the degree of matching between each trend feature and the corresponding standard behavior pattern and the possible deviation range.

[0115] S5028: Identify whether there are parameter combinations in the matching results that continuously deviate from the process behavior pattern. If so, generate an abnormal prompt message.

[0116] Specifically, the matching scores corresponding to the trend features included in the matching results undergo continuous analysis processing. Based on a preset deviation judgment threshold, a set of trend feature parameters with matching scores lower than the deviation judgment threshold is filtered. A time window sliding mechanism is used to identify trend feature parameter combinations that exhibit abnormal matching scores in multiple consecutive time periods. Combining the persistence characteristics and deviation magnitude characteristics of the matching score deviation, it is determined whether the trend feature parameter combination constitutes a behavior that continuously deviates from the process behavior pattern. When there is a trend feature parameter combination that meets the conditions for continuous deviation, the deviation behavior corresponding to the trend feature parameter combination is marked as abnormal, and an abnormality prompt message is generated. The abnormality prompt message includes the deviation parameter item, deviation start time, deviation end time, deviation degree, etc., which are used to describe the detailed characteristics of the continuous deviation process behavior pattern.

[0117] In one embodiment, such as Figure 6As shown, in step S60, if the quality inspection result meets the set quality index requirements, the packaging material information corresponding to the raw material combination information is called, and the packaged product after mixing is packaged to obtain the finished product information. This also includes:

[0118] S601: If the quality test results do not meet the quality index requirements, the target raw material information and processing node information associated with the mixed product are determined based on the traceability identification code, and the abnormal record information is obtained.

[0119] Specifically, the traceability identifier associated with the product after mixing is read, and the corresponding target raw material information, including raw material type, supply batch, acceptance time and pretreatment method, is retrieved from the raw material information database based on the traceability identifier. At the same time, the process node information involved in the process execution record database is located, and process data including weighing operation time, preparation ratio parameters, mixing processing time, and equipment operating status are obtained. The target raw material information and the process node information are structured and uniformly encoded, and combined to generate abnormal record information containing raw material batch and key process flow data.

[0120] S602: Generate corresponding feedback processing instructions based on the exception log information.

[0121] Specifically, based on the target raw material information and processing node information contained in the anomaly record information, key fields in the information are extracted as input for feedback analysis, including raw material batch number, processing time node, process parameter value, and anomaly occurrence stage. The correlation between each field is traversed and calculated through a preset causal analysis logic to identify raw materials or processing nodes with significant parameter shifts before and after the anomaly as dominant factors. The dominant factors are then matched with parameter anomaly response rules in the established process rule library to retrieve the processing strategy corresponding to the current dominant factor. The parameter items in the strategy template are then filled in with the current production status information, and finally, a feedback processing instruction with execution attributes is constructed.

[0122] Furthermore, the process rule base refers to a data set containing process control specifications, parameter adjustment rules, and anomaly handling strategies for each stage of flavor and fragrance preparation. The rules can be constructed based on historical experience, expert knowledge, or statistical optimization methods, and are used as a set of optimization instructions during feedback analysis.

[0123] In one embodiment, such as Figure 7 As shown, in step S602, a corresponding feedback processing instruction is generated based on the exception record information, including:

[0124] S6021: Use the causal source analysis mechanism to extract the target raw material information and processing node information from the abnormal record information to obtain the set of dominant factors.

[0125] In this embodiment, the causal tracing analysis mechanism refers to constructing a causal relationship graph based on the target raw material attributes and processing node parameters contained in the abnormal record information of the mixing process, and identifying a set of key factors with strong correlation to the abnormal behavior, which serves as the basis for generating feedback processing instructions.

[0126] Specifically, by constructing a causal tracing analysis mechanism based on time series and parameter change weights, the target raw material information and processing node information in the abnormal record information are converted into structured causal graph nodes, respectively. Key attribute items such as raw material category, supply batch, and physicochemical indicators are extracted from the raw material information, and the operation time, equipment status, and process parameter change values ​​are extracted from the processing node information. The correlation between the above factors is modeled based on Bayesian network or Granger causality determination method, the probability of each factor's influence on the final abnormal result is calculated, and factors with causal weight values ​​higher than a set threshold are selected as the dominant factor set.

[0127] S6022: Based on the set of dominant factors, retrieve matching optimization strategies from the process rule base and generate feedback processing instructions by combining them with the current production status parameters.

[0128] Specifically, each factor in the dominant factor set is input into the process rule base as a search condition. By matching feature fields such as raw material type, operation parameter offset direction and magnitude, and equipment operating status, the relevant historical optimization strategy records are called. Furthermore, based on the current production status parameters, such as real-time process ratio, current equipment load, and on-site personnel information, the matched optimization strategies are screened and adjusted. A weighted scoring mechanism is used to select the strategy template with the highest priority. The operation steps, adjustment parameters, and execution nodes in the strategy template are combined and encapsulated to finally generate a feedback processing instruction that includes the adjustment path, parameter correction value, and execution node time window.

[0129] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0130] In one embodiment, a device for testing and monitoring fragrances and flavors is provided, which corresponds one-to-one with a method for testing and monitoring fragrances and flavors in the above embodiments. For example... Figure 8As shown, this device for testing and monitoring flavors and fragrances includes a raw material binding module, a pretreatment module, a formulation preparation module, a mixing module, a quality inspection module, and a packaging execution module. Detailed descriptions of each functional module are as follows:

[0131] The raw material binding module is used to obtain the raw material information required for the preparation of flavorings and fragrances, to inspect the raw material information upon receipt, to generate a traceability identification code after the raw material information has passed inspection, and to establish a binding relationship between the traceability identification code and the raw material information to obtain the target raw material information.

[0132] The preprocessing module is used to perform preprocessing operations on the target raw material information to obtain preprocessed raw material information;

[0133] The preparation module is used to weigh the pretreated raw materials according to the formula requirements to obtain the weighed raw material group. The weighed raw material group is then used to perform the preparation operation to obtain the preparation raw material combination information.

[0134] The mixing module is used to perform mixing operations on the combination information of the raw materials to obtain the mixed product.

[0135] The quality inspection module is used to perform quality inspection on the mixed products and obtain the quality inspection results.

[0136] The packaging execution module is used to call the packaging material information corresponding to the raw material combination information if the quality inspection results meet the set quality index requirements, and perform packaging processing operations on the mixed product to obtain the finished product information.

[0137] Optionally, the raw material binding module includes:

[0138] The field extension submodule is used to perform field-level appending operations on the traceability identifier and raw material information, adding a field location in the raw material information to store the traceability identifier;

[0139] The binding registration submodule is used to write the traceability identification code into the field position and perform binding status registration. The raw material information that completes the binding relationship is the target raw material information.

[0140] Optional, the quality inspection module includes;

[0141] The parameter acquisition submodule is used to perform quality detection on the product after mixing and obtain mixing process parameter information;

[0142] The deviation identification submodule is used to input the mixing process parameter information into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing operation. If there is a process deviation in the mixing operation, an abnormal prompt message is generated.

[0143] Optionally, the deviation recognition submodule includes:

[0144] The historical parameter processing submodule is used to obtain historical mixing process parameter information and construct a multidimensional training sample set containing behavioral bias labels based on the historical mixing process parameter information;

[0145] The feature extraction submodule is used to input the multidimensional training sample set into the processing behavior monitoring model to be trained, and extract the time-series distribution features and fluctuation range features.

[0146] The feature discrimination construction submodule is used to construct a feature discrimination function to characterize the boundary of behavioral stability based on temporal distribution features and fluctuation range features;

[0147] The behavior classification and annotation submodule is used to classify and annotate the behavior deviation types of each sample in the multidimensional training sample set according to the feature discrimination function, and generate training supervision information.

[0148] The model training submodule is used to update the parameters of the processing behavior monitoring model to be trained by using a supervised learning algorithm based on the training supervision information, so as to obtain the trained processing behavior monitoring model.

[0149] Optional, the deviation recognition submodule includes:

[0150] The trend feature extraction unit is used to extract the changing trends of each parameter over time in the mixing process parameter information and construct a set of trend features.

[0151] The behavior pattern matching unit is used to match the trend feature set with the preset process behavior patterns in the trained processing behavior monitoring model to obtain the matching results;

[0152] An anomaly identification unit is used to identify whether there are parameter combinations in the matching results that continuously deviate from the process behavior pattern. If so, an anomaly prompt message is generated.

[0153] Optionally, the packaging execution module also includes:

[0154] The abnormal information backtracking submodule is used to determine the target raw material information and processing node information associated with the mixed product based on the traceability identification code if the quality test results do not meet the quality index requirements, and obtain abnormal record information.

[0155] The feedback instruction generation submodule is used to generate corresponding feedback processing instructions based on the exception log information.

[0156] Optionally, the feedback instruction generation submodule includes:

[0157] The dominant factor extraction unit is used to extract target raw material information and processing node information from abnormal record information using the causal tracing analysis mechanism, and obtain a set of dominant factors.

[0158] The optimization strategy generation unit is used to retrieve matching optimization strategies from the process rule base based on the set of dominant factors, and generate feedback processing instructions in combination with the current production status parameters.

[0159] For specific limitations regarding a device for testing and monitoring flavors and fragrances, please refer to the limitations regarding a method for testing and monitoring flavors and fragrances mentioned above, which will not be repeated here. The various modules in the aforementioned device for testing and monitoring flavors and fragrances can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0160] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for testing and monitoring fragrances and flavors.

[0161] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0162] Obtain information on raw materials required for the preparation of fragrances and flavors, conduct incoming inspection of raw materials, generate traceability identification codes after the raw materials pass inspection, and establish a binding relationship between the traceability identification codes and the raw material information to obtain the target raw material information;

[0163] Perform preprocessing operations on the target raw material information to obtain preprocessed raw material information;

[0164] Weigh the pretreated raw materials according to the formula requirements to obtain the weighed raw material group. Then, perform the preparation operation on the weighed raw material group to obtain the preparation raw material combination information.

[0165] Perform a mixing process on the raw material combination information to obtain the mixed product;

[0166] The product after mixing was subjected to quality testing to obtain the quality test results.

[0167] If the quality test results meet the set quality index requirements, the packaging material information corresponding to the raw material combination information is called, and the packaging process is performed on the mixed product to obtain the finished product information.

[0168] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0169] Obtain information on raw materials required for the preparation of fragrances and flavors, conduct incoming inspection of raw materials, generate traceability identification codes after the raw materials pass inspection, and establish a binding relationship between the traceability identification codes and the raw material information to obtain the target raw material information;

[0170] Perform preprocessing operations on the target raw material information to obtain preprocessed raw material information;

[0171] Weigh the pretreated raw materials according to the formula requirements to obtain the weighed raw material group. Then, perform the preparation operation on the weighed raw material group to obtain the preparation raw material combination information.

[0172] Perform a mixing process on the raw material combination information to obtain the mixed product;

[0173] The product after mixing was subjected to quality testing to obtain the quality test results.

[0174] If the quality test results meet the set quality index requirements, the packaging material information corresponding to the raw material combination information is called, and the packaging process is performed on the mixed product to obtain the finished product information.

[0175] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0176] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0177] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for monitoring the inspection of a fragrance, characterized in that, The method comprises: Obtaining raw material information required for essence and fragrance preparation, performing purchase inspection on the raw materials, generating a traceability identification code after the raw material inspection is qualified, and establishing a binding relationship between the traceability identification code and the raw material information to obtain target raw material information; According to the target raw material information, pre-treatment operation is performed on the raw materials to form pre-treated raw material information; According to the formula requirement, the pre-treated raw materials are weighed to obtain a plurality of groups of raw material quality data for generating corresponding weighed raw material group information, and the weighed raw materials are prepared to generate preparation raw material combination information; According to the preparation raw material combination information, the automatic mixing device performs mixing processing operation to obtain the product after mixing processing; Performing quality detection on the product after mixing processing to obtain quality detection results; If the quality detection results meet the set quality index requirement, the packaging material information corresponding to the preparation raw material combination information is called to perform packaging processing operation on the product after mixing processing, the unique identification code of each packaging unit is bound with the product traceability information, and finally the finished product data with complete information structure is formed as finished product information; The quality detection on the product after mixing processing to obtain quality detection results further comprises: Performing the quality detection on the product after mixing processing to obtain mixing process parameter information; The mixing process parameter information is input into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing processing operation, and if there is a process deviation in the mixing processing operation, an abnormal prompt information is generated; Before the mixing process parameter information is input into the trained processing behavior monitoring model, it further comprises: Obtaining historical mixing process parameter information, and constructing a multi-dimensional training sample set containing behavior deviation labels based on the historical mixing process parameter information; The multi-dimensional training sample set is input into the processing behavior monitoring model to be trained to extract time sequence distribution features and fluctuation interval features; Based on the time sequence distribution features and the fluctuation interval features, a feature discriminant function for describing the behavior stability boundary is constructed; According to the feature discriminant function, the behavior deviation type of each sample in the multi-dimensional training sample set is classified and labeled to generate training supervision information; Based on the training supervision information, a supervised learning algorithm is used to perform parameter update on the processing behavior monitoring model to be trained to obtain the trained processing behavior monitoring model; The mixing process parameter information is input into the trained processing behavior monitoring model to determine whether there is a process deviation in the mixing processing operation, and if there is a process deviation in the mixing processing operation, an abnormal prompt information is generated, which comprises: Extracting the change trend of each parameter in the mixing process parameter information over time to construct a trend feature set; The trend feature set is matched with the preset process behavior mode in the trained processing behavior monitoring model to obtain a matching result; Identify whether there is a parameter combination that continuously deviates from the process behavior mode in the matching result, and if there is, generate the abnormal prompt information.

2. A method for monitoring the quality of a fragrance according to claim 1, characterized in that, The binding relationship between the traceability identification code and the raw material information is established, and target raw material information is obtained, including: Performing a field level additional operation on the traceability identification code and the raw material information, adding a field position for storing the traceability identification code in the raw material information; Writing the traceability identification code into the field position and performing a binding state registration, and the raw material information after the binding relationship is completed is the target raw material information.

3. A method for monitoring the quality of a fragrance according to claim 1, characterized in that, The method further includes: If the quality detection result does not meet the quality index requirement, determining the target raw material information and processing process node information associated with the mixed processed product based on the traceability identification code, and obtaining abnormal record information; According to the abnormal record information, a corresponding feedback processing instruction is generated.

4. A method for monitoring the quality of a fragrance according to claim 3, characterized in that, According to the abnormal record information, a corresponding feedback processing instruction is generated, including: Using a cause and effect traceability analysis mechanism to extract the target raw material information and the processing process node information in the abnormal record information, and obtaining a dominant factor set; Based on the dominant factor set, an optimized strategy is retrieved in a process rule library, and the feedback processing instruction is generated in combination with the current production state parameters.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the method for inspecting and monitoring fragrance and flavor according to any one of claims 1 to 4.

6. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 5. The computer program is executed by the processor to realize the steps of the method for inspecting and monitoring fragrance and flavor according to any one of claims 1 to 4.

Images