Intelligent management and control system and method for food processing production line

Abstract

The invention provides an intelligent management and control system and method for a food processing production line, and the method comprises the steps: activating an equipment control algorithm to carry out the parameter adjustment of putting equipment if a difference characteristic vector exceeds a preset threshold range, determining the adjusted putting rate and volume parameters, and enabling the parameters to be optimized based on the equipment performance and material flow characteristics; according to the precision conformity mark, a formula proportion model of various raw materials is updated through a feedback mechanism, an updated formula proportion coefficient is determined, and the coefficient is used for adjusting the proportion relation between the raw materials to maintain formula balance; according to the stable fusion data set, a putting execution instruction sequence is generated, the instruction sequence comprises an action sequence and parameter configuration, and the instruction sequence is output to an actuator to achieve raw material putting control consistent in formula.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to an intelligent control system and method for a food processing production line. Background Technology

[0002] In the food processing industry, intelligent management of production lines has become a crucial pillar for improving product quality and production efficiency. As consumers' demands for food safety and quality continue to rise, ensuring precise operation and stable output during processing has become a key direction for industry development. The application of intelligent control systems is considered an important path to address the complexity and diverse needs of production; it not only concerns production efficiency but also directly impacts product taste and safety standards.

[0003] However, current management methods for many food processing production lines often face the dual challenges of adaptability and precision when dealing with complex formulas and diverse raw materials. Existing methods often struggle to flexibly respond to changes in the characteristics of different raw materials and processing conditions in dynamic production environments, especially when rapid switching of product types or adjustment of formula ratios is required. This can easily lead to operational deviations and inconsistent product quality. This limitation means that the control precision and response speed of the production process cannot meet the high standards required by the modern food industry. Focusing on the technical aspects, the core challenge of intelligent control of food processing production lines lies in achieving precise control over raw material input. Especially for materials in different forms, such as solid and liquid raw materials, the differences in their physical properties require the control system to have extremely high adaptability; otherwise, measurement deviations may occur, affecting the consistency of the final product's formula. A deeper issue is that for the addition of trace additives, due to their extremely small dosage but crucial effects, traditional measurement methods often cannot achieve the milligram-level precision required. This lack of precision directly manifests in actual production as an imbalance in the proportions of certain key components. For example, in processing a nutritional beverage, insufficient or excessive addition of micronutrients may lead to substandard product functional indicators and even negatively impact the consumer experience. Therefore, how to ensure high-precision control of the delivery of various materials in a dynamic production environment, and achieve extremely accurate measurement of trace components, has become a key problem that this study urgently needs to solve. Summary of the Invention

[0004] This invention provides an intelligent control system and method for a food processing production line, mainly comprising: The physical property data of solid and liquid materials in the production line are collected in real time by a sensor array. The physical property data includes key indicators such as density, viscosity and particle size distribution, and material property parameters are obtained to describe the basic properties of the materials. Based on the material characteristic parameters, a data analysis model is used to quantify the differences between solid and liquid materials to obtain a difference characteristic vector, which represents the deviation information of the two materials in physical properties. If the difference characteristic vector exceeds the preset threshold range, the equipment control algorithm is activated to adjust the parameters of the delivery equipment and determine the adjusted delivery rate and volume parameters. The parameters are optimized based on equipment performance and material flow characteristics. The metering requirement data of trace additives is extracted from the adjusted dosing rate and volume parameters. It is then determined whether the metering requirement data meets the preset high-precision standard to obtain a precision compliance mark. The mark reflects whether the metering precision meets the process requirements. For the accuracy compliance mark, the formula ratio model of various raw materials is updated through a feedback mechanism to determine the updated formula ratio coefficient. The coefficient is used to adjust the ratio relationship between raw materials to maintain formula balance. The updated formula ratio coefficient is obtained, and the dynamic environmental data of the production line is fused using an environmental data integration algorithm. The environmental data includes influencing factors such as temperature and humidity. The fused data is then judged to determine whether it has reached a stable state, and a stable fused dataset is obtained. Based on the stable fusion dataset, a sequence of delivery execution instructions is generated. The sequence of instructions includes the action order and parameter configuration, and is output to the actuator to achieve raw material delivery control with consistent formula.

[0005] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses an intelligent control system and method for food processing production lines. Addressing the challenges of significant differences in the properties of solid and liquid materials, the susceptibility of formulation ratios to environmental fluctuations, and the difficulty in ensuring accurate dosing in food processing, the system collects real-time data on the physical properties of materials, quantifies the vector of differences in characteristics, and adjusts the parameters of the dosing equipment when thresholds are exceeded to ensure optimal matching of dosing rate and volume. Simultaneously, it extracts the metering requirements data for trace additives, updates the formulation ratio coefficients based on accuracy compliance indicators, and integrates environmental data such as temperature and humidity to generate a stable dataset. Finally, it outputs a precise sequence of dosing execution instructions. This invention achieves dynamic adaptation of material properties to environmental factors through data analysis, feedback mechanisms, and environmental integration algorithms, solving the core challenges of formulation consistency and dosing accuracy, and significantly improving the intelligence level and processing quality stability of the production line. Attached Figure Description

[0006] Figure 1 This is a flowchart of an intelligent control system and method for a food processing production line according to the present invention.

[0007] Figure 2This is a schematic diagram of an intelligent control system and method for a food processing production line according to the present invention.

[0008] Figure 3 This is another schematic diagram of an intelligent control system and method for a food processing production line according to the present invention. Detailed Implementation

[0009] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0010] like Figures 1-3 This embodiment of an intelligent control system and method for a food processing production line may specifically include: S101. The physical property data of solid and liquid materials in the production line are collected in real time through a sensor array. The physical property data includes key indicators such as density, viscosity and particle size distribution, to obtain material property parameters that describe the basic properties of the materials.

[0011] Real-time data acquisition of solid and liquid materials in the production line is achieved through a sensor array, covering physical data such as density, viscosity, and particle distribution to obtain preliminary material characteristic parameters. Based on these parameters, a pre-established classification model is used to differentiate between solid and liquid materials, determining their respective physical data categories. If the classified physical data category differs from the preset material attribute standard, a data correction module adjusts the collected density and viscosity indices to obtain corrected material characteristic values. For the corrected material characteristic values, a support vector machine algorithm is used to perform pattern analysis on the particle distribution data to determine if the particle distribution meets a preset uniformity threshold. If the particle distribution does not meet the threshold, data smoothing is applied to optimize the distribution data, resulting in smoothed distribution parameters. Based on the smoothed distribution parameters and the corrected density and viscosity values, a comprehensive material attribute description is generated through a data integration module, determining the final basic material characterization result. Continuous monitoring of the comprehensive material attribute description updates the data parameters in the production line in real time, allowing for the acquisition of dynamically adjusted material characteristic states.

[0012] Specifically, in the automated production of high-end coatings or slurries, the sensor array serves as the core of data sensing, integrating a Coriolis mass flow meter for measuring density and viscosity, and a laser diffractometer for capturing particle distribution.

[0013] For example, when a production line conveys a mixed slurry, sensors collect its physical data in real time. If the collected density is 1.2 g / cm³, the viscosity is 500 cP, and the particle size is concentrated between 10 and 50 micrometers, this raw data constitutes the preliminary characteristic parameters of the material. The system uses a pre-trained classification model to determine whether the material is currently in a pure liquid solvent stage or a solid-liquid mixed suspension stage based on these parameters.

[0014] It is understandable that fluctuations in ambient temperature during actual operation may cause sensor drift, resulting in deviations between the classification results and the preset standards.

[0015] For example, if the model determines the current state is liquid, but the density reading rises abnormally due to a decrease in temperature, exceeding the preset range for liquid media, the data correction module intervenes. This module uses a thermodynamic compensation algorithm to adjust the collected density and viscosity indices, eliminate the influence of environmental noise, and obtain the corrected, true material characteristic values, ensuring the accuracy of the data's physical meaning.

[0016] In one possible implementation, a deep pattern analysis using a Support Vector Machine (SVM) algorithm is employed for the corrected data, particularly complex particle distribution data. The SVM constructs a high-dimensional hyperplane to compare the current particle distribution pattern with a standard uniform distribution model. If a bimodal distribution curve with high dispersion is detected, the SVM algorithm calculates the distance between its eigenvectors and a uniformity threshold. If the uniformity requirement is not met (e.g., agglomeration), an optimization process is triggered. At this point, data smoothing techniques, such as weighted moving averages, are used to filter out spikes and abrupt noise in the particle distribution curve, resulting in smoother distribution parameters that better reflect the overall trend of the material.

[0017] Preferably, the data integration module fuses the corrected density (e.g., corrected to 1.18 g / cm³), viscosity, and smoothed particle distribution parameters (e.g., D50 value stabilized at 25 micrometers) to generate a comprehensive material property description that includes rheological properties and microstructure. This description is not merely a static numerical record, but a dynamic digital fingerprint. Through continuous monitoring of this description, the production system can perceive changes in the material's state in real time. For example, when a change in raw material batches causes slight fluctuations in viscosity, the system can instantly update the characteristic state, providing precise decision-making basis for subsequent stirring speed or additive addition, thereby significantly improving product consistency and yield.

[0018] S102. Based on the material characteristic parameters, a data analysis model is used to quantify the differences between solid and liquid materials to obtain a difference characteristic vector, which represents the deviation information of the two materials in physical properties.

[0019] Physical property data of solid and liquid materials are acquired, focusing on key attributes such as density, viscosity, and particle size to obtain an initial attribute dataset. A pre-established data analysis model is used to process this initial attribute dataset, calculating the differences between solid and liquid materials in key attributes and determining an attribute difference matrix. Based on the attribute difference matrix, the main deviation information between solid and liquid materials is extracted, generating a difference characteristic vector reflecting the comparison results of the two materials in physical properties. Component analysis of the difference characteristic vector is performed to determine the deviation weight of each key attribute. If the deviation weight of an attribute exceeds a preset threshold, it is marked as a significantly different attribute, resulting in a set of significantly different attributes. For this set of significantly different attributes, the corresponding physical property data distribution is obtained, and a support vector machine model is used to classify the significantly different attributes, determining the classification boundary information. Based on the classification boundary information, a comparison mapping of significantly different attributes between solid and liquid materials is generated, yielding the final material comparison results for subsequent analysis.

[0020] In one embodiment, the physical properties of slurry and powder materials on the production line are acquired in real time using a high-precision sensor array.

[0021] For example, the collected density value for liquid slurry was 1.15 g / cm³, and the viscosity was 850 mPa·s, while for solid powder, the bulk density was 2.10 g / cm³, and the average particle size distribution was around 45 micrometers. These raw data constituted the initial attribute dataset, providing a foundation for subsequent quantitative analysis.

[0022] Understandably, by comparing the above dataset using a pre-defined data analysis model, the absolute differences between the two materials in various dimensions can be calculated.

[0023] For example, the difference in density is 0.95, while the difference in particle size represents the contrast between the continuity of the liquid phase and the discreteness of the solid phase. These differences are integrated into the property difference matrix, clearly depicting the physical gaps between materials.

[0024] In one embodiment, core deviation information is extracted from the difference matrix to generate a difference characteristic vector.

[0025] For example, density deviation, viscosity fluctuation, and particle distribution width can be transformed into coordinate points in a multi-dimensional vector space. By performing in-depth analysis of the components of this vector, the contribution of each attribute to material differentiation can be determined. If the weighted result of density deviation is 0.75, exceeding the preset threshold of 0.50, then this attribute is marked as a significantly differentiating attribute.

[0026] For example, for the set of marked significantly different attributes, the system will further retrieve its historical data distribution. Suppose historical data shows that the density of solid materials is typically concentrated in the range of 2.0 to 2.2, while that of liquid materials is concentrated in the range of 1.1 to 1.2. In this case, a support vector machine model is introduced to classify these significantly different attributes. This model finds an optimal hyperplane to partition the data points in the high-dimensional space.

[0027] Specifically, the model calculates the classification boundary information, which is the decision boundary that maximizes the distance between the two material types. In this example, the boundary might be set at a density of around 1.6. Materials falling on one side of the boundary are classified as liquid, and those on the other side as solid. This classification is not based on a single numerical value but rather a comprehensive boundary incorporating multi-dimensional data such as viscosity and particle size. Based on the determined classification boundary information, the system generates a comparative mapping of significantly different attributes between solid and liquid materials. This is like constructing a multi-dimensional coordinate system, projecting the current real-time data points into it, and visually displaying the degree of deviation from the standard solid-liquid state. The final material comparison results are no longer simple numerical lists but structured information after weighting and classification. For example, it may explicitly indicate that the current batch of liquid material has shifted 15% towards solid characteristics in terms of viscosity, suggesting a potential risk of solidification. This approach effectively solves the ambiguity of single-indicator judgments. Through multi-attribute fusion and difference amplification, it ensures the accuracy of material state identification, providing a reliable quantitative basis for subsequent production process adjustments.

[0028] S103. If the difference characteristic vector exceeds the preset threshold range, the equipment control algorithm is activated to adjust the parameters of the delivery equipment and determine the adjusted delivery rate and volume parameters. The parameters are optimized based on equipment performance and material flow characteristics.

[0029] Step 1: Acquire real-time operating data from the dispensing equipment, compare and analyze the differences in characteristics and vector ranges. If a difference exceeds a preset threshold, record the current data status and obtain an anomaly trigger signal. Step 2: Based on the anomaly trigger signal, activate the equipment control algorithm, perform comprehensive calculations on equipment performance and material flow data, and determine the adjusted dispensing rate parameters. Step 3: Using the dispensing rate parameters and material flow characteristics, perform secondary calibration on the volume parameters to obtain optimized volume parameter values. Step 4: Using the optimized volume parameter values ​​and dispensing rate parameters, issue adjustment commands to the equipment control module and determine if the equipment has completed parameter updates. Step 5: If the equipment has completed parameter updates, collect the updated operating data, monitor the material flow status in real time, and obtain flow stability indicators. Step 6: Based on the flow stability indicators, analyze the matching degree between equipment performance and dispensing rate to determine if it meets preset operating standards. Step 7: If the matching degree meets the operating standards, save the current parameter configuration, continuously monitor subsequent difference data, and determine the stability of equipment operation.

[0030] For example, in a solid-liquid material mixing and dispensing scenario, when acquiring real-time operating data of the dispensing equipment, the system continuously monitors the change in drag torque during the mixing process of solid particles and liquid solvent. When the viscosity deviation component in the previously quantified difference characteristic vector exceeds a preset threshold of 0.8, it indicates that the current liquid material cannot effectively encapsulate the solid particles. The system immediately records the data of this high-resistance state, thereby obtaining an anomaly trigger signal. This real-time comparison mechanism can accurately capture the operational risks caused by sudden changes in the physical properties of materials.

[0031] In one possible implementation, upon receiving an abnormal trigger signal, the equipment control algorithm is quickly activated. This algorithm reads the current output power of the drive motor and the flow rate of the material in the pipeline for comprehensive calculation. If the particle size of the solid material is large, causing overall slow flow, the algorithm will reduce the original dispensing rate from 50 kg / min to 35 kg / min, determining the adjusted dispensing rate parameter, effectively avoiding equipment overload and pipeline blockage.

[0032] It should be noted that the secondary calibration of the volume parameters is based on the actual flow characteristics of the material. Due to the wall adhesion effect of liquid materials under different operating conditions, the system will adjust the volume parameters corresponding to the original valve opening based on the adjusted dispensing rate. Assuming that the original valve opening corresponds to a theoretical volume of 10 liters, under the current high viscosity flow characteristics, the actual effective throughput may only be 8 liters. Therefore, the system will compensate the target value of the volume parameters to 12.5 liters to obtain the optimized volume parameter value, ensuring the accuracy of the material ratio.

[0033] For example, after the optimized volume parameter values ​​and the new dispensing rate are sent to the control module, the system collects updated sensor data. By analyzing the variance of flow fluctuations over a continuous operating cycle, a flow stability index is obtained. If this variance is less than 0.05, it is determined that the equipment performance is highly matched with the current dispensing rate and meets the preset operating standards. Subsequently, the system saves the current parameter configuration and continuously monitors subsequent differential characteristic data, ensuring extremely high stability during the long-term dispensing process of solid and liquid materials.

[0034] S104. Extract the metering requirement data of trace additives from the adjusted dosing rate and volume parameters, determine whether the metering requirement data meets the preset high precision standard, and obtain the precision compliance mark. The mark reflects whether the metering precision meets the process requirements.

[0035] Raw measurement data of trace additives are obtained from records of dosing rate and volume parameters. Outliers are removed through data cleaning to obtain a preliminary measurement dataset. This dataset is then compared and analyzed against preset standards. Data values ​​deviating from these standards are marked as non-compliant, resulting in a categorized dataset. The non-compliant data is extracted and its degree of deviation is determined through statistical analysis, yielding deviation analysis results. Based on these results and the high precision requirements of the preset standards, deviations exceeding preset thresholds are flagged as inaccuracy, resulting in a precision judgment flag. This flag is then used to correlate with process specifications, and logical comparison analysis determines whether the specifications are met, yielding process compliance results. These compliance results are then stored in a record-keeping manner, linked to the raw data of dosing rate and volume parameters, resulting in the final analysis record. Finally, a data visualization tool is used to generate a comparison chart of measurement accuracy and process compliance, determining the overall execution status of the business process.

[0036] Specifically, in scenarios involving the precise delivery of trace additives, obtaining raw measurement data is fundamental to quality control. When the system extracts data from historical records of delivery rates and volume parameters, it often contains transient noise caused by sensor jitter or signal transmission delays. By cleaning the data and removing these outliers that significantly deviate from physical probability, the system can reconstruct the actual operating trajectory of the equipment and obtain a highly reliable preliminary measurement dataset. This process not only eliminates interference but also provides a clean data source for subsequent refined analysis, effectively preventing misjudgments.

[0037] Understandably, the system uses preset standards to perform point-by-point comparison and analysis on the processed dataset.

[0038] For example, if the standard dispensing rate is set at 5.2 L / h, with an allowable fluctuation range of ±0.1 L / h, and the monitored value remains consistently at 5.4 L / h for a certain period, this data segment is marked as unacceptable. In this case, the system does not simply discard the data; instead, it extracts these unacceptable portions for in-depth statistical analysis, calculating the degree of deviation. If the calculated deviation reaches 3.8%, while the preset high-precision threshold is only 2.0%, the system will immediately generate a flag indicating inaccuracy. This tiered judgment strategy can keenly detect minute measurement deviations that may affect the final product performance, ensuring the rigor of dispensing accuracy.

[0039] In one embodiment, based on the generated accuracy judgment flag, the system further performs logical verification by associating it with process specification requirements. Assume the process specification explicitly states that under specific temperature and pressure conditions, the cumulative error in the dosage volume must not exceed a specific limit. The system compares the accuracy flag with these complex logical conditions. If it finds that although the single-point rate deviation is small, the cumulative volume parameter has led to a proportioning imbalance, it determines that the specification conditions are not met, and outputs a result indicating whether the process complies with the specifications. Finally, these analysis results are associated with and saved with the original dosage rate and volume parameters, and comparison charts are generated using visualization tools.

[0040] For example, the chart clearly displays the curve of measurement accuracy changing over time and the status color blocks of process compliance, helping managers to intuitively determine the execution status of business processes, thereby achieving comprehensive traceability and optimization of the deployment process.

[0041] S105. For the accuracy compliance mark, update the formula ratio model of various raw materials through a feedback mechanism to determine the updated formula ratio coefficient. The coefficient is used to adjust the ratio relationship between raw materials to maintain the formula balance.

[0042] By collecting historical data and real-time feedback information, the status of the accuracy indicator is analyzed to obtain preliminary deviation assessment results. Based on the deviation assessment results, a pre-established formula ratio model is used to calculate the adjustment direction of each raw material ratio and determine the preliminary update coefficient range. If the preliminary update coefficient range exceeds a preset threshold, the raw material ratio relationship is recalibrated through the information processing stage to obtain calibrated coefficient data. Based on the calibrated coefficient data and the real-time data stream of the feedback mechanism, the balance status of each raw material formula is judged to determine the correction requirements for formula balance. Based on the correction requirements, a logistic regression model is used to iteratively update the ratio model to determine the final formula ratio coefficients. Based on the final formula ratio coefficients, the ratio relationship between raw materials is adjusted to obtain an updated formula balance scheme. If the updated formula balance scheme deviates from the requirements of the accuracy indicator, the data from the feedback mechanism is analyzed in depth through the information processing stage to obtain new adjustment basis.

[0043] Specifically, in the scenario of controlling the proportion of trace additives, the system first determines the initial state based on the accuracy compliance indicator generated in the previous step. Suppose historical data shows that while the dosing rate of a key catalyst is within the allowable range, it consistently hovers near the lower limit, causing the accuracy indicator to frequently display a critical state. In this case, by analyzing the deviation assessment results of this state and combining them with a preset formulation ratio model, the system calculates that the ratio of the catalyst to the main reaction liquid needs fine-tuning. For example, the initial update coefficient indicates that the dosing amount should be increased by 1.5%. If this initial update coefficient exceeds the system's preset safety threshold of 1.2%, the system will not directly apply this parameter but will instead perform a secondary calibration of the raw material ratio through information processing. This process aims to eliminate misjudgments caused by instantaneous sensor noise or pipeline pressure fluctuations, ensuring the reliability of the adjustment basis.

[0044] For example, during the calibration phase, the system, combined with real-time data streams from the feedback mechanism, discovered that although the flow rate was low, the volume parameters within the reactor remained stable, indicating that the raw material formulations were in a dynamic equilibrium. Blindly increasing the catalyst too much could disrupt this equilibrium. Based on this correction requirement, the system used a logistic regression model to iteratively update the original proportioning model. By introducing multi-dimensional variables such as volume, flow rate, and historical accuracy indicators, the model underwent multiple iterations, correcting the initially aggressive 1.5% increase coefficient to a smoother 0.8%, and determining the final formulation proportion coefficient. This approach not only achieved precise adjustment of the proportions between raw materials and obtained an updated formulation balance scheme, but also effectively avoided process oscillations caused by over-adjustment. If the updated scheme still deviates from the high-precision requirements of the accuracy indicators during simulation or trial operation, the system will trigger a deep analysis mechanism to uncover deeper influencing factors such as raw material purity fluctuations or pump wear from the feedback data, providing new adjustment basis for the next round of precise control, thereby ensuring that the metering of trace additives always meets stringent process specifications.

[0045] S106. Obtain the updated formula ratio coefficient, use an environmental data integration algorithm to fuse the dynamic environmental data of the production line, the environmental data including influencing factors such as temperature and humidity, determine whether the fused data has reached a stable state, and obtain a stable fused dataset.

[0046] Dynamic environmental data, including raw records of temperature and humidity factors, is acquired through a real-time acquisition module on the production line and stored in a pre-defined data warehouse to obtain a preliminary environmental data set. Based on this preliminary set, an environmental data integration algorithm is used to standardize the temperature and humidity factors, generating a processed environmental feature dataset. For this processed dataset, a pre-defined judgment standard is applied to analyze data fluctuation ranges. If the fluctuation range exceeds a pre-defined threshold, outliers are smoothed to obtain a smoothed environmental feature dataset. Key environmental variables are extracted from the smoothed dataset and, combined with the initial values ​​of the formula ratios, adjustment coefficients are calculated to determine the adjusted formula ratio dataset. The adjusted formula ratio dataset is then acquired and analyzed for its compatibility with a stable state. If the compatibility is lower than a pre-defined threshold, the adjustment coefficients are iteratively optimized to obtain an optimized formula ratio dataset. Using the optimized dataset, a final fused dataset is generated. Its compatibility with the stable state requirements is assessed, and a stable fused dataset that meets the standards is output. This stable fused dataset is stored in a pre-defined database and transmitted to the production line control module via a data interface, completing the data closed-loop processing.

[0047] Specifically, in the processing of dynamic environmental data from the production line, temperature and humidity are key variables affecting chemical reaction rates and the physical properties of raw materials, and the accuracy of their data directly determines the effectiveness of formula adjustments. Through the real-time acquisition module of the production line, the system can capture environmental parameters at millisecond-level frequencies. For example, in a certain acquisition cycle, the system recorded a workshop temperature of 28.5 degrees Celsius and a relative humidity of 62%. To eliminate calculation errors caused by different units of measurement, the environmental data integration algorithm standardizes these raw records, mapping them to a standard range of 0 to 1, constructing an environmental feature dataset with a unified scale. This process not only improves the convergence speed of subsequent model calculations but also ensures the comparability of multi-source heterogeneous data.

[0048] For example, in the analysis of data fluctuation range, the system presets strict stability thresholds. Suppose that due to electromagnetic interference, the sensor's temperature reading momentarily spikes to 50 degrees Celsius, which clearly exceeds normal physical change logic. In this case, the judgment criteria will identify this anomaly and use smoothing techniques, such as moving averages, to correct the anomaly to a trend-compliant 28.6 degrees Celsius, referencing normal values ​​from preceding and following periods. This smoothing of anomalies effectively eliminates environmental noise, prevents misjudgments of formula proportions due to data fluctuations, and thus ensures the continuity and safety of the production process.

[0049] In one possible implementation, extracting key environmental variables and calculating adjustment coefficients is the core of achieving precise ingredient proportioning. The system analyzes the impact of environmental characteristics on the activity of specific raw materials, based on the initial values ​​of the formula proportions. If the current ambient temperature is too high, it may lead to faster volatilization of raw material A. Based on this, the system calculates that the formula proportion of raw material A needs to be increased by an adjustment coefficient of 0.05. Subsequently, the system simulates the reaction state of the adjusted formula in a virtual environment. If the analyzed steady-state matching degree is only 85%, lower than the preset 90% threshold, an iterative optimization mechanism is triggered. The system fine-tunes the coefficient to 0.052 and verifies again until the matching degree reaches the target. This iterative optimization based on the dynamic environment ensures that the final generated fusion dataset can offset environmental interference in real time, keeping the instructions transmitted to the control module through the data interface within the optimal process window, thus achieving closed-loop control from environmental perception to decision execution.

[0050] S107. Based on the stable fusion dataset, generate a sequence of delivery execution instructions, which includes the action sequence and parameter configuration, and output it to the actuator to achieve consistent raw material delivery control.

[0051] By extracting key information from stable and fused datasets, an initial data processing framework is constructed to obtain the raw data set for subsequent instruction generation. Based on the raw data set, the sequence of raw material dispensing actions is analyzed, and a pre-defined rule base is used for matching to determine the time node and execution priority for each action. If the time node and execution priority meet a pre-defined threshold range, corresponding execution instructions are generated, and the instruction sequence related to the action sequence is obtained. For the generated instruction sequence, combined with the constraints of the parameter configuration, a logic verification tool is used to check whether the parameter configuration meets the requirement of formula consistency. If the parameter configuration meets the requirement of formula consistency, the instruction sequence and parameter configuration are integrated into a complete dispensing control scheme and transmitted to the actuator. The actuator receives the dispensing control scheme, parses the execution instructions and parameter configuration, and obtains the final control implementation path. Based on the control implementation path, the status changes of raw material dispensing are monitored in real time, key data during the execution process is recorded, and the complete closed loop of dispensing control is determined.

[0052] Specifically, building the initial data processing framework involves transforming the stable, fused dataset, corrected for environmental factors in the preceding steps, into concrete physical execution logic. This process is not merely a simple data transfer, but rather a mapping of abstract formula proportion coefficients into underlying signals recognizable by production line actuators.

[0053] For example, when determining the order of raw material dispensing, the system analyzes the material characteristics in the original data set. Assuming the production scenario involves mixing powdered main materials and liquid additives, the preset rule base includes an anti-caking dispensing strategy, requiring the powder layer to be established before the liquid additive is dispensed. Accordingly, the system sets the execution priority of the powder dispensing action to level one, with a time node marked as 0 seconds; while the priority of the liquid additive dispensing action is set to level two, with a time node set as 15 seconds. If the system detects that the overlap between the liquid additive preparation command time and the powder dispensing end time exceeds a threshold of 2 seconds, it determines it as a potential conflict and automatically adjusts the time nodes to meet process requirements.

[0054] In one possible implementation, the logic verification tool performs a deep check on the generated instruction sequence. When the instruction sequence requires the dispensing of hygroscopic raw materials at maximum flow rate under specific humidity conditions, the verification tool will make a judgment based on the environmental flow rate constraint table in the parameter configuration. If the current instruction sets the flow rate to 50 liters per minute, while the constraint condition states that the flow rate should not exceed 40 liters per minute under high humidity to avoid pipe blockage, then the parameter configuration is determined to not meet the formulation consistency requirements, and a correction process needs to be triggered.

[0055] Preferably, when the actuator analyzes the control implementation path, it establishes a millisecond-level real-time monitoring mechanism. When the actuator opens the valve according to the plan, if there is a deviation between the actual value fed back by the flow sensor and the command setting value, the system will record this key data in real time and dynamically fine-tune the duration of subsequent actions, thereby eliminating the cumulative error caused by environmental fluctuations or equipment aging, ensuring the accuracy of each batch of raw material feeding, and realizing a tight closed loop from data fusion to physical execution.

[0056] If the technical solution of this application involves the acquisition of personal information, the product using this solution has clearly informed the user of the processing rules and obtained the user's consent before processing. If sensitive personal information is involved, the user's individual consent has been obtained and the "express consent" requirement has been met. For example, a clear sign is placed at the collection device to indicate the collection scope, and the user's voluntary entry is considered as consent; or authorization is obtained through pop-up windows, user uploads, etc. The processing rules include the processor, purpose, method, and type of information.

[0057] 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.

Claims

1. An intelligent control system and method for a food processing production line, characterized in that, The method includes: The physical property data of solid and liquid materials in the production line are collected in real time by a sensor array. The physical property data includes key indicators such as density, viscosity and particle size distribution, and material property parameters are obtained to describe the basic properties of the materials. Based on the material characteristic parameters, a data analysis model is used to quantify the differences between solid and liquid materials to obtain a difference characteristic vector, which represents the deviation information of the two materials in physical properties. If the difference characteristic vector exceeds the preset threshold range, the equipment control algorithm is activated to adjust the parameters of the delivery equipment and determine the adjusted delivery rate and volume parameters. The parameters are optimized based on equipment performance and material flow characteristics. The metering requirement data of trace additives is extracted from the adjusted dosing rate and volume parameters. It is then determined whether the metering requirement data meets the preset high-precision standard to obtain a precision compliance mark. The mark reflects whether the metering precision meets the process requirements. For the accuracy compliance mark, the formula ratio model of various raw materials is updated through a feedback mechanism to determine the updated formula ratio coefficient. The coefficient is used to adjust the ratio relationship between raw materials to maintain formula balance. The updated formula ratio coefficient is obtained, and the dynamic environmental data of the production line is fused using an environmental data integration algorithm. The environmental data includes influencing factors such as temperature and humidity. The fused data is then judged to determine whether it has reached a stable state, and a stable fused dataset is obtained. Based on the stable fusion dataset, a sequence of delivery execution instructions is generated. The sequence of instructions includes the action order and parameter configuration, and is output to the actuator to achieve raw material delivery control with consistent formula.

2. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, The process involves real-time acquisition of physical property data of solid and liquid materials in the production line using a sensor array. This physical property data includes key indicators such as density, viscosity, and particle size distribution, yielding material characteristic parameters that describe the basic properties of the materials, including: By using a sensor array to collect real-time data on solid and liquid materials in the production line, covering physical data such as density, viscosity and particle distribution, preliminary material characteristic parameters can be obtained. Based on the collected material characteristic parameters, a pre-established classification model is used to distinguish between solid and liquid materials and determine their respective physical data categories. If the category of the classified physical data is inconsistent with the preset material property standard, the data correction module will adjust the collected density and viscosity indices to obtain the corrected material characteristic values. Based on the corrected material characteristic values, a support vector machine algorithm is used to perform pattern analysis on the particle distribution data to determine whether the particle distribution meets the preset uniformity threshold. If the particle distribution does not reach the preset uniformity threshold, the particle distribution data is optimized and adjusted through data smoothing to obtain smoothed distribution parameters. Based on the smoothed distribution parameters, combined with the correction values ​​of density and viscosity indices, a comprehensive material property description is generated through the data integration module to determine the final basic material characterization results. By continuously monitoring the comprehensive material attribute description, the data parameters in the production line are updated in real time to obtain the dynamically adjusted material characteristic status.

3. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, The step involves quantifying the differences between solid and liquid materials using a data analysis model based on the material characteristic parameters, resulting in a difference characteristic vector. This vector represents the deviation information in physical properties between the two materials, including: Acquire physical property data of solid and liquid materials, and collect data on key properties such as density, viscosity and particle size to obtain an initial property dataset; Using a pre-established data analysis model, the initial attribute dataset is processed to calculate the differences between solid and liquid materials in key attributes and determine the attribute difference matrix. Based on the attribute difference matrix, the main deviation information between solid materials and liquid materials is extracted to generate a difference characteristic vector, which reflects the comparison results of the physical properties of the two materials. By analyzing the components of the difference characteristic vector, the deviation weight of each key attribute is determined. If the deviation weight of a certain attribute exceeds a preset threshold, it is marked as a significantly different attribute, thus obtaining a set of significantly different attributes. For a set of significantly different attributes, obtain the corresponding physical attribute data distribution, use a support vector machine model to classify the significantly different attributes, and determine the classification boundary information; Based on the classification boundary information, a comparative mapping of the significantly different attributes of solid and liquid materials is generated to obtain the final material comparison results for subsequent analysis and processing.

4. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, If the difference characteristic vector exceeds a preset threshold range, the equipment control algorithm is activated to adjust the parameters of the delivery equipment, determining the adjusted delivery rate and volume parameters. These parameters are optimized based on equipment performance and material flow characteristics, including: Step 1: Obtain real-time operating data of the delivery device, compare and analyze the difference characteristics and vector range, and if the difference characteristics exceed the preset threshold, record the current data status and obtain the abnormal trigger signal. Step 2: Based on the abnormal trigger signal, activate the equipment control algorithm, perform comprehensive calculations on equipment performance and material flow data, and determine the adjusted delivery rate parameters; Step 3: By combining the delivery rate parameter with the material flow characteristics, the volume parameter is calibrated a second time to obtain the optimized volume parameter value; Step 4: Using the optimized volume parameter values ​​and delivery rate parameters, issue adjustment instructions to the equipment control module to determine whether the equipment has completed the parameter update; Step 5: If the equipment completes the parameter update, collect the updated operating data, monitor the material flow status in real time, and obtain the flow stability index. Step Six: Based on the flow stability index, analyze the matching degree between equipment performance and delivery rate to determine whether the preset operating standards are met; Step 7: If the matching degree meets the operating standards, save the current parameter configuration and continuously monitor subsequent difference characteristic data to determine the stability of equipment operation.

5. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, The process involves extracting the metering requirement data for trace additives from the adjusted dosing rate and volume parameters, determining whether the metering requirement data meets a preset high-precision standard, and obtaining a precision compliance indicator. This indicator reflects whether the metering precision meets the process requirements, including: Raw measurement data of trace additives were obtained from records of dosing rate and volume parameters. Outliers were removed through data cleaning to obtain a preliminary measurement dataset. For the initially compiled measurement dataset, a comparative analysis is performed using preset standards. If the measurement values ​​in the dataset deviate from the range of the preset standards, they are marked as unqualified data, resulting in a classified dataset. Based on the classified dataset, extract the data portion marked as unqualified, determine its degree of deviation through statistical analysis, and obtain the deviation analysis results; Based on the deviation analysis results and the high precision requirements of the preset standards, if the deviation exceeds the preset threshold, a flag indicating that the precision does not meet the requirements is generated, thus obtaining a precision judgment flag. Based on the accuracy judgment criteria and the requirements of the associated process specifications, the process compliance results are obtained by logical comparison and analysis to determine whether the specification conditions are met. For the process compliance results, a record storage method is adopted to link and save the results with the original data of the dosing rate and volume parameters to obtain the final analysis record; Based on the final analysis records, comparative charts of measurement accuracy and process compliance are generated using data visualization tools to determine the execution status of the overall business process.

6. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, Regarding the accuracy compliance indicator, the formula ratio model for various raw materials is updated through a feedback mechanism to determine the updated formula ratio coefficients. These coefficients are used to adjust the proportional relationships between raw materials to maintain formula balance, including: By collecting historical data and real-time feedback information, the status of the accuracy indicator is analyzed to obtain preliminary deviation assessment results; Based on the deviation assessment results, a pre-established formula ratio model is used to calculate the adjustment direction of each raw material ratio and determine the preliminary update coefficient range. If the initial update coefficient range exceeds the preset threshold, the raw material ratio relationship will be calibrated a second time through the information processing stage to obtain the calibrated coefficient data. Based on the calibrated coefficient data and combined with the real-time data stream of the feedback mechanism, the balance state of each raw material formula is determined, and the correction requirements for the formula balance are derived. By revising the requirements, the proportion model is iteratively updated using a logistic regression model to determine the final formula proportion coefficients. Based on the final formula ratio coefficient, adjust the proportions between raw materials to obtain an updated formula balance scheme; If the updated formula balance scheme deviates from the accuracy mark requirements, the data from the feedback mechanism will be analyzed in depth through the information processing stage to obtain new adjustment basis.

7. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, The updated formula ratio coefficient is obtained by using an environmental data integration algorithm to fuse dynamic environmental data from the production line. This environmental data includes influencing factors such as temperature and humidity. The algorithm then determines whether the fused data has reached a stable state, resulting in a stable fused dataset, including: Dynamic environmental data is acquired through a real-time acquisition module on the production line. The data includes raw records of temperature and humidity factors and is stored in a preset data warehouse to obtain a preliminary set of environmental data. Based on the preliminary environmental dataset, an environmental data integration algorithm was used to standardize the temperature and humidity factors, generating a processed environmental feature dataset. For the processed environmental feature dataset, a preset judgment standard is invoked to analyze the data fluctuation range. If the fluctuation range exceeds the preset threshold, the outliers are smoothed to obtain a smoothed environmental feature dataset. Key environmental variables are extracted from the smoothed environmental feature dataset. Combined with the initial values ​​of the formula ratio, the adjustment coefficients of the environmental variables on the formula ratio are calculated to determine the adjusted formula ratio dataset. Obtain the adjusted formula ratio dataset, analyze its matching degree with the steady state, and if the matching degree is lower than the preset threshold, iteratively optimize the adjustment coefficient to obtain the optimized formula ratio dataset. The final fusion dataset is generated using the optimized formula ratio dataset. It is then determined whether the dataset meets the requirements for a stable state, and a stable fusion dataset that meets the standards is output. For stable fusion datasets, they are stored in a preset database and transmitted to the production line control module through a data interface to complete data closed-loop processing.

8. The intelligent control system and method for a food processing production line according to claim 1, characterized in that, The step involves generating a sequence of delivery execution instructions based on the stable fusion dataset. This sequence includes an action order and parameter configuration, which is then output to the actuator to achieve consistent raw material delivery control. This includes: By extracting key information from stable and fused datasets, an initial data processing framework is constructed to obtain the raw data set for subsequent instruction generation. Based on the original data set, the sequence of actions for raw material delivery is analyzed, and a pre-set rule base is used for matching to determine the time node and execution priority of each action. If the time node and execution priority meet the preset threshold range, the corresponding execution instruction is generated, and the instruction sequence content related to the action order is obtained; For the generated instruction sequence, combined with the constraints of the parameter configuration, a logic verification tool is used to check whether the parameter configuration meets the requirement of recipe consistency. If the parameter configuration meets the requirement of recipe consistency, the instruction sequence and parameter configuration will be integrated into a complete dispensing control scheme and transmitted to the actuator. The actuator receives the control scheme, parses the execution instructions and parameter configurations, and obtains the final control implementation path. Based on the control implementation path, monitor the changes in the status of raw material delivery in real time, record key data during the execution process, and determine the complete closed loop of delivery control.

Images