Extrusion processing digital twin construction method and system based on flow resistance state

Abstract

This invention discloses a method and system for constructing a digital twin of extrusion processing based on flow resistance state. The method collects equipment operating parameters, calculates the flow resistance state quantity and the exponentially smoothed flow resistance state quantity, monitors them until a stable operating condition is reached, obtains the baseline flow resistance state quantity under stable operating conditions, performs normalization processing to construct the relative flow resistance state quantity and multi-timescale trend state, comprehensively judges the multi-timescale trend state to generate process suggestion information and constructs an equipment state table; establishes an HDMR model to determine the optimal adjustment scheme, and then performs feedback adjustment based on the optimal adjustment scheme and the deviation between the current relative flow resistance state quantity and the target relative flow resistance state quantity to calculate the current control quantity; constrains the current control quantity and generates control instructions for equipment operating parameters; after receiving the control instructions, the PLC adjusts the physical control components to achieve real-time deviation correction.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing and industrial digital twin technology in food processing. Specifically, it relates to a method and system for constructing a digital twin of extrusion processing based on flow resistance state, and in particular, a method and system for constructing the state of extrusion processing based on the fusion of Venturi pressure difference, multi-timescale data and multi-source operating condition data. It is applicable to the monitoring of the operating status and process optimization of continuous processing of rice products, grain products and other extruded foods. Background Technology

[0002] Extrusion processing is a commonly used continuous processing technology in the food industry, and it has been widely applied in the large-scale production of rice products, puffed foods, and pet foods. During extrusion processing, materials undergo complex rheological and structural changes under the action of screw extrusion, shearing, and heat. The process state is affected by a variety of factors, including feed rate, moisture content, temperature, and mechanical energy.

[0003] However, existing extrusion process monitoring methods have significant limitations. Currently, industrial sites mainly rely on experience to judge conventional equipment parameters (such as current, temperature and speed) and the state of the extruded material, which makes it difficult to accurately and directly reflect the true flow state of the material in the extrusion chamber.

[0004] To obtain rheological information about materials, traditional online rheological measurement methods are typically based on the principles of capillary rheometers or slit rheometers. They calculate shear rate and shear stress by measuring the material's pressure difference, flow rate, and channel geometry parameters, thereby obtaining the absolute viscosity. However, these methods face the following technical bottlenecks in practical industrial applications:

[0005] 1. This method usually requires additional configuration of multiple detection devices such as flow meters and density sensors, resulting in a complex system structure and high cost.

[0006] 2. Because the inside of the extrusion equipment is in an extreme environment of high temperature, high pressure and high shear coupling for a long time, and the food material system usually exhibits non-Newtonian fluid characteristics, multiphase mixing, easy adhesion and easy accumulation, the relevant online rheological measurement devices face problems such as limited installation space and insufficient long-term stable operation when directly integrated into the limited internal space of the extrusion equipment.

[0007] Given the aforementioned limitations of hardware testing, existing factories still rely on offline testing to determine the rheological state of materials during extrusion processing. This method struggles to achieve real-time online characterization and feedback. Furthermore, in actual production, when blockages or process abnormalities occur during extrusion, operators often need to check the equipment status one by one to determine the source of the problem, sometimes even requiring shutdown and disassembly for cleaning. This not only results in significant material and energy consumption but also leads to a decrease in production efficiency.

[0008] Therefore, without changing the existing mechanical structure of the extrusion equipment or relying on the absolute physical properties of the material, how to utilize existing industrial sensor data to construct an interpretable state model that reflects the internal state of the material during the extrusion process, and on this basis, realize real-time monitoring and intelligent control of the extrusion process, has become an important technical issue for promoting the intelligent upgrading of food extrusion processing. Summary of the Invention

[0009] The purpose of this invention is to address the shortcomings of existing technologies by providing a method and system for constructing a digital twin of extrusion processing based on flow resistance state. Without the need to obtain material properties such as absolute viscosity or density, and without the need to install an additional viscosity meter, this invention utilizes only the pressure sensor data from the industrial production line to construct a digital twin state model that reflects the internal flow and structural evolution state of the extrusion process, thereby achieving interpretable modeling and process optimization of the extrusion process.

[0010] The objective of this invention is achieved through the following technical solution:

[0011] In a first aspect, the present invention provides a method for constructing a digital twin of an extrusion process based on flow resistance state, comprising the following steps:

[0012] S1: Collect equipment operating parameters to form continuous time series data, and obtain the set of operating batch time windows based on the judgment rules;

[0013] S2: Collect equipment operating parameters in real time within the time window of the running batch to obtain the batch-level running dataset of the corresponding batch;

[0014] S3: Calculate the flow resistance state variables The flow resistance state variables are obtained by outlier detection, filtering, and exponential smoothing. ; and then regarding the above or Monitoring continues until a steady-state condition is reached, under which the baseline flow resistance state quantity is obtained. ;

[0015] S4: Based on the above , , Normalization is performed to construct the relative flow resistance state quantity. ;

[0016] S5: Based on the above Construct trend states across multiple time scales;

[0017] S6: Perform a comprehensive judgment on the multi-timescale trend status to generate process suggestion information and construct an equipment status table;

[0018] S7 establishes an HDMR model and sets the target relative flow resistance state quantity, determines the optimal adjustment scheme, and then uses the equipment operating parameter group corresponding to the optimal adjustment scheme as the benchmark adjustment scheme for the current control input. It also performs feedback adjustment based on the deviation between the current relative flow resistance state quantity and the target relative flow resistance state quantity, and calculates the current control quantity.

[0019] S8: The current control quantity is constrained to obtain the constrained control quantity and control instructions for the equipment operation parameters are generated. After receiving the control instructions, the PLC can adjust the physical control components to achieve real-time deviation correction.

[0020] Furthermore, in step S1, the equipment operating parameters include the operating status of the feeding motor, the venturi inlet pressure, the venturi outlet pressure, the feeding flow rate, the water flow rate of the conditioner, the steam flow rate of the conditioner, and the mechanical energy.

[0021] The specific determination rule is as follows: if equipment shutdown is detected, the downtime is accumulated; if the equipment starts running again and the accumulated downtime is greater than or equal to a preset threshold, a new time window is determined to begin, and a unique identifier ID is generated; at the same time as the new batch begins, the previous window is marked as completed, and the accumulated output, flow rate, and operating status parameters of the current experimental window are initialized; when the equipment changes from running state to stopped state, or the duration of the feed flow interruption exceeds the preset threshold, the current running batch is determined to end.

[0022] Further, in step S3,

[0023] The flow resistance state quantity is specifically the pressure difference between the Venturi inlet pressure and the Venturi outlet pressure.

[0024] The stable operating condition specifically refers to: when or Within the preset time window The fluctuation range within is less than the threshold. The moment when this condition is first met is recorded as the candidate start time. The preset time window A continuous time interval is used to determine the stability of the flow resistance state quantity; then a stability determination process is performed, wherein the stability criterion for stability determination is: if from the candidate start time... Initially, the state quantity of flow resistance or Stability determination time The threshold is always met. The stability determination time If the time range is 20–60 s, the system is determined to have entered a stable operating condition, and the candidate start time is recorded. Backtracking confirmed as a stable start time If at the stability determination time If the stability criterion is not met at any given time, then the candidate start time is determined. If the system fails, a new candidate start time is searched, and the above stability determination process is repeated until the system enters a stable operating condition; wherein, the threshold... The range is 0.02–0.05 bar;

[0025] The reference flow resistance state quantity Specifically: Regarding The number of sampling points is n Take the arithmetic mean, where, This is the baseline calculation time.

[0026] Further, in step S4,

[0027] The normalization process specifically involves: if ,but For As molecules, with Fractions with a denominator of 1 are not 1; otherwise, they are written as 1. ;in, This is an invalid value.

[0028] Further, in step S5,

[0029] The multi-timescale trend states include short-term relative flow resistance state quantities and long-term relative flow resistance state quantities.

[0030] The short-term relative flow resistance state quantity Specifically: The difference; where, This indicates a short duration, ranging from 30 to 60 seconds. for The relative flow resistance state quantity before;

[0031] The long-term relative flow resistance state quantity Specifically: The difference; where, Indicates a long duration, ranging from 240 to 360 seconds. for The relative flow resistance state quantity before.

[0032] Further, in step S6,

[0033] The comprehensive judgment refers to:

[0034] For the long term, if If the output process suggestion is: check the water supply to the equipment, and suggest reducing the water supply by 1-3%; if If the output process suggestion is: check the water addition adjustment, it is recommended to increase the water addition by 1-3%, and check the plasticization state of the material;

[0035] Among them, the As the lower limit of the long-term trend change, the This represents the upper limit of the long-term trend change.

[0036] In the short term, if Output process suggestion information: Check feeding stability; the aforementioned This is the lower limit of short-term trend change; if Output process suggestion information: Check feeding stability; the aforementioned This represents the upper limit of short-term trend changes;

[0037] like and All are within the range of variation. Output process recommendation information: Maintain the current operating status and keep the original operating conditions.

[0038] The equipment status table includes timestamps, relative flow resistance status values, multi-timescale trend status, equipment motor operating status, feed flow rate, water addition variation, steam flow rate, and unit energy consumption index.

[0039] Further, in step S7,

[0040] The establishment of the HDMR model specifically involves: The constant term of the HDMR model Individual device operating parameters right Impact Backup operating parameters and The interaction of Impact higher-order interaction pairs The sum of the effects of γ; where, The equipment operating parameter variables input to the HDMR model, i.e., the training samples X

[0041] The target relative flow resistance state quantity is used to characterize the desired extrusion process state;

[0042] The determination of the optimal mediation scheme specifically involves:

[0043] Calculate the predicted relative flow resistance state quantity corresponding to the k-th candidate adjustment scheme. : Set the candidate equipment operating parameter group corresponding to the k-th candidate adjustment scheme Substituting into the HDMR model, the candidate adjustment scheme is obtained by setting the adjustment range;

[0044] Then compare the predicted relative flow resistance state quantities corresponding to the candidate adjustment schemes. Based on the deviation from the target relative flow resistance state quantity, the candidate adjustment scheme that makes the predicted relative flow resistance state quantity closest to the target state is selected as the optimal adjustment scheme.

[0045] Further, in step S8,

[0046] The control instructions for generating equipment operating parameters based on the constrained processing quantities specifically involve mapping the constrained control quantities into physical quantity instructions recognizable by the PLC. This mapping process includes: determining the physical quantity form and unit based on the actuator type corresponding to each control variable, including feed flow rate (t / h), water addition change (t / h), and steam flow rate (kg / h); proportionally converting the control quantities according to the actuator's range to meet the PLC's setpoint range; discretizing the control quantities based on the actuator's minimum adjustment step size; performing necessary smoothing or speed limiting processing on the control signals based on the actuator's response characteristics; and encapsulating the processed control quantities into PLC register or variable formats and sending them to the PLC system for execution via an industrial communication protocol.

[0047] The physical control components include a water inlet valve and a feeding motor.

[0048] Secondly, the present invention also provides a digital twin system for food extrusion processing using the aforementioned method, comprising:

[0049] The data acquisition module is used to collect equipment operating parameters;

[0050] The data transmission module is used to transmit the equipment operating parameters collected by the data acquisition module to form a raw data table;

[0051] The data processing module processes the original data table and divides it into running batches to obtain a set of running batch time windows, and calculates the flow resistance state quantity, the baseline flow resistance state quantity, the relative flow resistance state quantity, and the short-term and long-term relative flow resistance state quantities.

[0052] The equipment status construction module receives the equipment operating parameters, relative flow resistance state quantities, and multi-timescale trend analysis results output by the data processing module, integrates them, and constructs an equipment status table.

[0053] The process decision module receives multi-timescale trend analysis results from the equipment status construction module and generates process recommendation information.

[0054] The blockage alarm analysis module receives the operating status of the feeding motor and the feeding flow rate from the equipment status construction module, performs anomaly identification, and outputs blockage warning or blockage alarm information when the threshold conditions are met, and records the alarm time and duration.

[0055] The front-end visualization module receives data from the data processing module, the equipment status construction module, and the congestion alarm analysis module, and displays it through the digital twin platform interface.

[0056] Furthermore, the data acquisition module includes a feeding system, a conditioning system, a Venturi detection unit, an extrusion processing system, and a control and data processing system. The feeding system includes a feeding hopper, a feeding motor, and a flow sensor. The conditioning system includes a water addition unit, a steam heating unit, a temperature sensor, and a flow sensor. The Venturi detection unit includes a pressure sensor. The flow sensor, temperature sensor, and pressure sensor are specifically industrial sensors used to collect equipment operating parameters. The control and data processing system includes a PLC control unit and an edge computing unit used to mark the collected equipment operating parameters with a unified timestamp.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] (1) The present invention constructs the flow resistance state quantity based on the Venturi pressure difference and further constructs the relative flow resistance state quantity. It can effectively characterize the flow state of the material during the extrusion process without obtaining the absolute physical property parameters of the material, thus reducing the dependence of model building on complex experimental measurements.

[0059] (2) By utilizing the existing pressure sensor of the extrusion equipment to achieve state construction, there is no need to introduce an additional online rheological testing device, which significantly improves the feasibility and engineering adaptability of the method in industrial production lines;

[0060] (3) The present invention constructs a state characterization system with relative flow resistance state quantity as the core, which establishes a direct correlation between the operating parameters of the extrusion equipment and the material flow behavior, and improves the interpretability and perceptibility of the process state.

[0061] (4) By introducing a multi-timescale analysis mechanism that combines short-term and long-term analysis, the influence of instantaneous fluctuations on state judgment is effectively suppressed, and the stability and robustness of state recognition are improved.

[0062] (5) Based on the state variables, a mapping relationship between equipment operating parameters and process state is constructed, and combined with the optimal adjustment scheme and error feedback adjustment, dynamic closed-loop control of the extrusion process is realized, which improves the process control accuracy;

[0063] (6) The method of the present invention can be integrated with digital twin systems and process optimization systems to achieve integrated state perception, predictive analysis and real-time control of the extrusion process, thereby improving the overall intelligence level of the system. Compared with traditional control methods based on empirical parameters or single process variables, the control mechanism of the present invention based on state variable driving can more accurately reflect the changes in material rheological characteristics, thereby improving the control adaptability under complex working conditions. Attached Figure Description

[0064] Figure 1 This is a flowchart of the extrusion process;

[0065] Figure 2 A schematic diagram of the steps in a digital twin system for extrusion processing;

[0066] Figure 3 A schematic diagram illustrating the flow resistance state under Venturi pressure differential;

[0067] Figure 4 A schematic diagram illustrating the multi-timescale trend analysis of relative flow resistance state quantities;

[0068] Figure 5 This is a schematic diagram of the structure of a digital twin system for the extrusion process. Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] It should be noted that, unless otherwise specified, the features in the following embodiments and implementation methods can be combined with each other.

[0071] In a first aspect, the present invention provides a method for constructing a digital twin of extrusion processing based on the characterization of flow resistance state, such as... Figure 1 , 2 As shown, the method specifically includes the following steps:

[0072] S0 Batch Time Window Recognition

[0073] This step arranges the equipment operating parameters generated by the PLC system of the extrusion processing equipment according to a unified timestamp order to form continuous time series data D(t). Based on the judgment rules, it is segmented to obtain the set of operating batch time windows W. ᵢ And select the current running batch time window W ᵢ As the object of subsequent data processing.

[0074] The equipment operating parameters include the feed motor running status (FeedMotorRunning, 1 / 0, running / stopping) and the Venturi inlet pressure. (Unit: Bar) Venturi export pressure The data includes (unit: Bar), Feed Flow (unit: t / h), Conditioner Water Flow (unit: kg / h), Conditioner Steam Rate (unit: kg / h), and Mechanical Energy (unit: kW·h / t). The set of operating batch time windows is used to identify each continuous production run segment and serves as the basis for subsequent digital twin analysis and closed-loop control, ensuring the continuity and accuracy of the experimental data. Each collected equipment operating parameter has a unified timestamp to ensure data synchronization and traceability.

[0075] The determination rules are as follows:

[0076] (1) If a device stoppage is detected (FeedMotorRunning=0), then start accumulating the stoppage time;

[0077] (2) If the device runs again (FeedMotorRunning=1) and the cumulative downtime is ≥ the preset threshold STOP_WINDOW_SEC (set to 30 s), it is determined that a new time window has started and a unique identifier ID is generated. The identifier ID preferably adopts the timestamp format (YYYY-MM-DD HH:mm:ss).

[0078] (3) At the same time as the new batch starts, mark the previous window as completed (Status='FINISHED'), and initialize the cumulative output, flow rate and running status parameters of the current experimental window;

[0079] (4) When the equipment changes from running state to stopped state, or the feeding flow interruption lasts for more than the preset threshold, the current running batch is determined to be over.

[0080] Based on the above judgment rules, the continuous time series data D(t) is segmented to obtain multiple running batch time windows Wᵢ. The data in each time window are then classified according to the corresponding batch identifier ID, so that the data in the same time window belong to the same running batch.

[0081] Through the above processing, the running batch time window Wᵢ is used to limit the scope of subsequent data and serves as a unified data boundary for flow resistance modeling, multi-timescale analysis, blockage alarms, and process recommendations.

[0082] S1 Data Acquisition

[0083] In the running batch time window W ᵢ Internally, the system collects equipment operating parameters in real time through industrial sensors and a PLC control system to obtain batch-level operating datasets D for each batch. ᵢ The collected equipment operating parameters are marked with a unified timestamp and categorized according to the time window Wᵢ of their respective operating batches to form a batch-level dataset Dᵢ. This dataset is then input into a local database for data caching and management, and then synchronously transmitted to the factory edge database via an independent industrial area network to form a raw data table for subsequent state calculations and process optimization of the digital twin model.

[0084] S2 Flow Resistance State Construction

[0085] S2-1 Calculation of Flow Resistance State Quantities

[0086] During extrusion processing, the material is continuously subjected to shearing, compression, and heat under the conveying and plasticizing action of the screw. Its internal state (such as moisture content and starch gelatinization degree) constantly changes, leading to variations in resistance during flow. When the material is forced to flow through the Venturi structure (Venturi tube) within the extrusion chamber, the contraction and expansion of the flow channel cross-section cause local velocity changes, resulting in a pressure difference between the inlet and outlet of the Venturi tube. Therefore, this method utilizes the Venturi pressure difference as an important parameter characterizing the internal state of the extrusion process, treating it as a flow resistance state quantity within the Venturi structure. This flow resistance state quantity characterizes the changes in flow resistance of the extruded material within the equipment, including but not limited to flow resistance, structural density, and plasticization state.

[0087] Specifically, such as Figure 3 As shown, the Venturi inlet pressure in the operational data set Dᵢ obtained in step S1 is... (Unit: Bar) and Venturi outlet pressure (Unit: Bar) The pressure difference is used as the flow resistance state quantity. :

[0088]

[0089] in, For a moment Venturi inlet pressure, For a moment Venturi export pressure.

[0090] The flow resistance state quantity It not only directly reflects the flow resistance of raw materials within the expansion chamber, but also serves as a fundamental indicator for subsequent multi-timescale rheological analysis, process recommendations, and blockage alarms. To ensure data integrity and continuity, It is stored in real time in the database view v_venturi_delta_p.

[0091] Furthermore, regarding Outlier detection and filtering are performed to remove abnormal data caused by instantaneous fluctuations or sensors.

[0092] Set upper and lower limits for differential pressure and The upper and lower limits of the differential pressure are preferably set based on historical batch data and equipment experience. When in real-time... or When an outlier is detected, it is either discarded or marked as NULL and not used in subsequent calculations. Based on this, the sliding average flow resistance state quantity is calculated for n consecutive sampling points. :

[0093]

[0094] Where n is the sliding window size, i.e. the number of sampling points, in this example n=5; i is the summation index, representing the time step of backtracking; It is the flow resistance state quantity at the i-th sampling point forward from the current time t.

[0095] For the sliding average flow resistance state quantity Further exponential smoothing is performed to obtain the exponentially smoothed flow resistance state quantity. :

[0096]

[0097] in, Smoothing coefficient ( ) This is the exponentially smoothed flow resistance state quantity from the previous moment.

[0098] Finally, the exponentially smoothed flow resistance state quantity can be obtained. This is used for subsequent analysis.

[0099] S2-2 Calculation of Reference Flow Resistance State Quantities

[0100] Within each equipment operation batch time window, after system startup, the flow resistance state quantity is... or Continuous monitoring is performed when the flow resistance state quantity or Within the preset time window The fluctuation range within is less than the threshold. The threshold The range is 0.02 to 0.05 bar, and the moment when this condition is first met is recorded as the candidate start time. The preset time window A continuous time interval is used to determine the stability of the flow resistance state quantity. Then, a stability determination process is performed, where the stability criterion is: if the stability is determined from the candidate start time... Initially, the state quantity of flow resistance or Stability determination time The threshold is always met. The stability determination time If the time range is 20–60 s, the system is determined to have entered a stable operating condition, and the candidate start time is recorded. Backtracking confirmed as a stable start time If at the stability determination time If the stability criterion is not met at any given time, then the candidate start time is determined. If the system fails, a new candidate start time is searched, and the stability determination process is repeated until the system enters a stable operating condition.

[0101] Under the aforementioned stable operating conditions, the arithmetic mean of the Venturi pressure difference signal is taken to obtain the baseline flow resistance state quantity. :

[0102]

[0103] in, Time interval The i-th sampling time within the range, where n is the number of sampling points. This is the baseline calculation time, ranging from 30 to 120 seconds; It is a moment The state quantity of flow resistance.

[0104] S3 Calculation of relative flow resistance state quantities

[0105] In actual industrial production, there are differences in the types of materials, formulas and operating conditions during the extrusion process. Therefore, it is difficult to compare the states between different operating conditions by directly using the pressure difference value.

[0106] Therefore, this method is based on the flow resistance state quantity obtained in step S2. Flow resistance state quantity after exponential smoothing Reference flow resistance state quantity By introducing a baseline flow resistance value and normalizing it, the relative flow resistance state quantity is obtained. :

[0107]

[0108] in, For the current moment The relative flow resistance state quantity, The baseline flow resistance value, This is an invalid value.

[0109] The above normalization process eliminates the influence of differences in absolute pressure values ​​under different operating conditions, thus reducing the relative flow resistance state quantity. It can reflect the relative changing trend of the internal rheological state during the extrusion process, providing a basic indicator for subsequent multi-timescale analysis, process recommendation, and blockage alarm.

[0110] S4 Multi-Timescale Trend Analysis

[0111] During extrusion processing, the operating status of the equipment may be affected by instantaneous fluctuations, such as changes in feed rate, water addition adjustments, or raw material fluctuations. Judging solely based on instantaneous conditions can easily lead to misjudgments. Therefore, this invention introduces a multi-timescale synchronous analysis method, calculating the trend of state changes through different time windows to improve the stability of state identification.

[0112] Specifically, to reduce the impact of instantaneous operating conditions on state judgment, such as Figure 4 The calculation of relative flow resistance state quantities is shown. The short-term and long-term changes.

[0113] (1) Short-term ( Relative flow resistance state quantity :

[0114]

[0115] in, The range is 30-60 s, the for The relative flow resistance state quantity before.

[0116] The short-term ( Relative pressure difference change It reflects the short-term, instantaneous material flow within the expansion chamber, and is used to identify short-term operating condition fluctuations.

[0117] (2) Long-term ( Relative flow resistance state quantity :

[0118]

[0119] in, This is a state quantity representing relative flow resistance. The range is 240-360 s, the for The relative flow resistance state quantity before.

[0120] The long-term ( Relative flow resistance state quantity Reflected in The pressure difference change trend is used to identify process trend changes over a longer period of time, such as the material gradually becoming thicker or thinner, and to determine the overall process deviation.

[0121] This step, through multi-timescale trend analysis, can more stably identify state changes during the extrusion process. These state changes are specifically the relative flow resistance state quantities at different time scales, including short-term and long-term relative flow resistance state quantities. By synchronizing the above multi-timescale data, a stable extrusion process state index is constructed.

[0122] S5 outputs process recommendations and builds equipment status tables.

[0123] S5-1 Output Process Recommendation Information

[0124] Combined with the multi-timescale trend state obtained in step S4 ( , The system comprehensively assesses the data to generate corresponding process recommendations. These recommendations can be output in real time via a digital twin system, providing operators with a reference for process adjustments and restoring material rheology to an ideal state. When both short-term and long-term changes exceed the threshold, the long-term trend takes priority; if the long-term trend is stable while short-term fluctuations exceed the threshold, the operator is alerted but not directly intervened, achieving a balance between safety and flexibility.

[0125] Specifically, the comprehensive judgment refers to:

[0126] For long-term trends ( ) Changes, if they occur This indicates that the material has become thinner over a long period, resulting in a decrease in viscosity. The system generates an intervention command and outputs process recommendations: check the water addition level in the equipment and suggest reducing the water addition by 1-3%. This represents the lower limit of long-term trend changes, ranging from -0.12 to -0.08.

[0127] Conversely, when This indicates that the material has been thickening for a long time, the pressure difference has been rising continuously, and plasticization may have occurred. The system generates an intervention command and outputs process suggestion information: check the water addition adjustment, it is recommended to increase the water addition by 1-3%, and check the plasticization status of the material. This represents the upper limit of the long-term trend change, ranging from 0.08 to 0.12.

[0128] For short-term trend (60 s) changes, the operator is prompted to monitor the equipment, but no direct intervention is triggered. This indicates that the material has become thinner in a short period of time, requiring the operator to pay attention to the water volume and the stability of the feeding. The system outputs process suggestion information: Check the feeding stability. This represents the lower limit of short-term trend changes, ranging from -0.06 to -0.04.

[0129] when This indicates that the material has thickened instantaneously, requiring the operator to monitor pressure stability. The system outputs a process suggestion: check feeding stability. This represents the upper limit of short-term trend changes, ranging from 0.04 to 0.06.

[0130] if and If all conditions are within the threshold, the system outputs the following process suggestion: Hold current conditions and maintain the original operating conditions.

[0131] S5-2 Constructing the Equipment Status Table

[0132] By integrating equipment operating parameters, relative flow resistance state quantities, and multi-timescale trend quantities, a digital twin equipment state table is constructed, specifically including timestamps (YYYY-MM-DD HH:mm:ss), relative flow resistance state quantities (... Multi-timescale trend status ( , The status table displays the equipment's motor running status (FeedMotorRunning), feed flow rate (t / h), water addition variation (kg / h), steam flow rate (kg / h), and unit energy consumption (kW.h / t). This status table is used for front-end trend chart display and also provides input data for blockage alarms, HDMR predictive model construction, and closed-loop control.

[0133] S6 HDMR Prediction and Closed-Loop Control

[0134] S6-1 Building and Training an HDMR Model

[0135] This step further removes outliers and missing values ​​from the historical batch data of equipment operating parameters and relative flow resistance state quantities in the equipment status table obtained in S5-2, and constructs training samples X (equipment operating parameters) and y (relative flow resistance state quantities). ) dataset.

[0136] First, the mapping relationship between equipment operating parameters and relative flow resistance state variables is decomposed into a tractable low-dimensional function combination to reduce computational complexity. An HDMR model is then established to calculate and predict the relative flow resistance state variables.

[0137]

[0138] in, To predict the relative flow resistance state quantity, For the constant term of the HDMR model, Operating parameters for a single device right The impact, For equipment operating parameters and The interaction of The impact, The equipment operating parameter variables input to the HDMR model are the training samples X; To include higher-order interaction pairs The impact.

[0139] Historical batch data is input into the HDMR model for training. Cross-validation is used to evaluate the model accuracy, and the optimal model parameters are determined by combining the mean squared error (MSE). The predicted relative flow resistance state variables are then output. .

[0140] During real-time operation, the target relative flow resistance state quantity is set as follows: The target relative flow resistance state quantity It can be determined based on the baseline state under stable operating conditions, the historical best operating state, or the preset process target, and is used to characterize the desired extrusion process state. Further, the target relative flow resistance state quantity... It can also be preset or adaptively updated according to different materials to adapt to the process requirements under different production conditions.

[0141] S6-2 Determine the optimal adjustment scheme

[0142] Furthermore, under the current equipment operating conditions, several preset adjustment ranges are set for equipment operating parameters such as water supply, feed flow rate, and steam flow rate, forming multiple candidate adjustment schemes. These candidate adjustment schemes are specifically candidate equipment operating parameter groups. Let the candidate equipment operating parameter group corresponding to the k-th candidate adjustment scheme be denoted as... .

[0143] Based on the candidate adjustment schemes, the operating parameter groups of each candidate device are input into the HDMR prediction model, and the predicted relative flow resistance state quantity corresponding to the k-th candidate adjustment scheme is calculated respectively. :

[0144]

[0145] Based on the predicted relative flow resistance state quantities corresponding to each candidate adjustment scheme, compare them with the target relative flow resistance state quantities. The deviation is considered, and the optimal adjustment scheme is selected from multiple candidate adjustment schemes as the current control strategy, taking into account the magnitude of control quantity changes, the degree of process fluctuations, and equipment operating constraints. Preferably, the candidate adjustment scheme that makes the predicted relative flow resistance state quantity closest to the target state can be selected as the optimal adjustment scheme, i.e.:

[0146]

[0147] Where arg is an abbreviation for argument (independent variable); arg min represents the number of the adjustment scheme that minimizes the objective function among all candidate adjustment schemes; This is the number corresponding to the optimal candidate adjustment scheme.

[0148] S6-3 Calculation of Control Variables and Constraint Handling

[0149] After determining the optimal adjustment scheme k*, set its corresponding equipment operating parameters. This serves as the baseline adjustment scheme for the current control input. Based on this, the error between the current real-time relative flow resistance state quantity and the target relative flow resistance state quantity is further calculated using S3. :

[0150]

[0151] in, Real-time relative flow resistance state quantities calculated based on Venturi pressure difference.

[0152] Then, a proportional-integral controller is used to reduce the error. Adjustment increment converted into control variable :

[0153]

[0154] in, This is the proportionality coefficient. is the integral coefficient.

[0155] Adjusting the increment Based on this, a fuzzy neural network (FNN) is preferably added for compensation to correct the control deficiencies of the proportional-integral controller under nonlinear operating conditions or model errors. The compensation amount... The mathematical expression is:

[0156]

[0157] Among them, the The error between the target relative flow resistance and the actual relative flow resistance state quantity; The error change rate; These are the equipment operating parameters; This is the amount of compensation.

[0158] This leads to the total control quantity. for:

[0159]

[0160] Then update the control variables based on the total control quantity, and the current control quantity. for:

[0161]

[0162] Among them, the The control quantity of the previous moment, the This is the total control quantity.

[0163] The The current control variable represents the control command used to adjust the operating parameters of the equipment. Includes one or more of the following sub-control quantities: feed flow rate Used to control the feeding motor and adjust the feeding flow rate (t / h); water flow rate in the conditioner. Used to control the water supply rate (t / h) of the conditioner water pump; and the steam flow rate of the conditioner. It is used to control equipment operating parameters such as steam supply (kg / h).

[0164] S7 Control Execution

[0165] Before control execution, the current control quantity is... Constraint processing includes (1) setting upper and lower limits on control variables to meet the operating range of the extrusion equipment; (2) quantifying the control quantity according to the minimum adjustment step of the actuator to match the actual execution accuracy; and (3) for multivariate control scenarios, weighting the control quantity based on the sensitivity of each control variable to the relative flow resistance state quantity and the current deviation to determine the final adjustment amount of each control variable.

[0166] After the above processing, the constrained control quantity is obtained, and control instructions for the equipment operating parameters are generated based on the control quantity.

[0167] The digital twin system maps the constrained control quantities into physical quantity commands that the PLC can recognize. The mapping process includes:

[0168] (1) Determine the physical quantity form and unit according to the actuator type corresponding to each control variable, including feed flow rate (t / h), water addition change (t / h) and steam flow rate (kg / h).

[0169] (2) Based on the range of the actuator, the control quantity is proportionally converted to meet the range of the PLC setting value;

[0170] (3) Discretize the control quantity according to the minimum adjustment step size of the actuator;

[0171] (4) Combine the actuator response characteristics (such as delay and inertia) to perform necessary smoothing or speed limiting processing on the control signal;

[0172] (5) Encapsulate the processed control quantity into a PLC register or variable format and send it to the PLC system for execution via industrial communication protocol.

[0173] After receiving the control command, the PLC adjusts the physical control components such as the water valve and the feeding motor according to the set value to complete the operation of the physical environment and realize real-time deviation correction.

[0174] During execution, sensors collect real-time data on equipment operating status, transmitting data such as Venturi inlet pressure, Venturi outlet pressure, feed flow rate, water addition variation, and steam flow rate to the digital twin system. This updates the status table and re-executes candidate regulation scheme prediction, optimal regulation scheme selection, and control quantity correction, thus forming a closed-loop control system. With the accumulation of production data, preferably, the system can continuously optimize HDMR prediction model parameters and control parameters. , A closed-loop control strategy is employed to improve the stability and control robustness of the extrusion process under different material conditions.

[0175] The method also includes a congestion alarm:

[0176] One of the common problems during the operation of extrusion equipment is that materials are prone to blockage inside the equipment, which hinders material transport, increases equipment load, and often requires shutdown to clear the blockage, seriously affecting the continuity of production.

[0177] To enable early warning of congestion risks, this step involves congestion alarm determination. By monitoring the equipment's operating status and flow rate trends in real time, potential congestion situations are identified and assessed.

[0178] The blockage alarm determination is based on the device motor running status (FeedMotorRunning) and feeding flow rate in the device status table obtained in step S5-2. When the device is in running status (FeedMotorRunning=1) and the feeding flow rate is lower than the second preset threshold WATCH_TH (1.1 to 1.5 times FLOW_TH), the system determines that there is an abnormal event candidate interval. The abnormal event candidate interval specifically represents the possible material conveying abnormality or blockage risk.

[0179] When the equipment is in operation (FeedMotorRunning=1) and the feed flow rate is lower than the first preset threshold FLOW_TH (range 0.5-2 L / h), the system determines that there is an abnormal event candidate interval, which specifically represents the potential risk of blockage. After determining the abnormal event candidate interval, the system continuously counts the number of time points that meet the conditions (alarm_count), and records the first occurrence time (alarm_start) and the last occurrence time (alarm_end) of the abnormal event to describe the duration range of the abnormal event. Then, the alarm event and related data are written into the alarm data table and returned to the front-end system in the form of API through the digital twin platform interface to realize the visualization and decision support of potential blockage risks. The alarm event and related data include: (1) the first occurrence time (alarm_start) and the last occurrence time (alarm_end) of the abnormal state; (2) the number of time points that meet the conditions (alarm_count); (3) alarm type information, which is determined to be a blockage alarm based on the relationship between the feed flow rate and the preset threshold.

[0180] Compared with existing technologies, this method does not require absolute material properties and can construct the extrusion process state using only the existing pressure sensors in the equipment; it has strong engineering adaptability, does not require the addition of a complex online rheometer, and can be directly applied to industrial production lines; it constructs an interpretable state model, reflecting changes in material flow behavior through relative flow resistance state quantities; it improves the stability of state identification through multi-timescale synchronization; and it can be integrated with digital twin systems, intelligent control systems, and process optimization systems.

[0181] Secondly, this invention provides a digital twin system for extrusion processing based on flow resistance state characterization, such as... Figure 5 As shown, it includes the following modules:

[0182] (1) Data acquisition module: including feeding system, conditioning system, venturi detection unit, extrusion processing system, control and data processing system; used to collect the operating parameters of the extrusion equipment, including the feed motor running (FeedMotorRunning, 1 / 0, running / stopping), venturi inlet pressure (Unit: Bar) Venturi export pressure The parameters are as follows: (unit: Bar), Feed Flow (unit: t / h), Conditioner Water Flow (unit: kg / h), Conditioner Steam Rate (unit: kg / h), and Mechanical Energy Menergy (unit: kW·h / t). These equipment operating parameters are specifically collected by industrial sensors.

[0183] The industrial sensors are arranged in different functional units of the extrusion processing system, specifically including:

[0184] The feeding system includes a feeding hopper, a feeding motor, and a flow sensor. The feeding hopper is used to store and buffer the material to be extruded. The feeding motor is used to drive the material from the feeding hopper to the extruder inlet and output the operating status of the feeding motor (1 / 0, running / stopping). The flow sensor is used to measure the feeding flow rate (unit: t / h) in real time.

[0185] The conditioning system is equipped with a water supply unit and a steam heating unit, as well as a temperature sensor and a flow sensor. The water supply unit is used to supply water to the material, the steam heating unit is used to heat and condition the material, the temperature sensor is used to monitor the material temperature, and the flow sensor is used to measure the water flow rate (unit kg / h) and the steam flow rate (unit kg / h) of the conditioner.

[0186] The Venturi detection unit is located inside the expansion chamber of the twin-screw extruder, and includes pressure sensors located at the converging section and the necking outlet of the Venturi tube (attached). Figure 2 ), used to obtain Venturi inlet pressure in real time Export pressure A temperature sensor can be selected as needed to assist in analyzing the effect of material temperature on rheological behavior. The pressure sensor is a high-temperature and high-pressure resistant industrial pressure sensor, and its output signal is synchronously acquired by the data acquisition module and marked with a unified timestamp.

[0187] The control and data processing system includes a PLC control unit and an edge computing unit, which are used to collect equipment operating parameters from various industrial sensors, mark the collected equipment operating parameters with a unified timestamp, input them into a local database for data caching and management, and then transmit them synchronously to the factory edge database through an independent industrial area network for flow resistance state construction, multi-timescale trend analysis, process suggestion information output, equipment status table construction, HDMR prediction and closed-loop control.

[0188] (2) Data transmission module: This module transmits the equipment operating parameters collected by the data acquisition module to the factory edge database via an independent industrial area network to form raw data tables. To prevent access to industrial control and production data by the office system or external network and the Internet, and to ensure the data security of the digital twin system, the independent industrial area network is physically or logically isolated from the office network and the Internet to ensure the real-time performance, security, and integrity of production data, and to meet the requirements of the digital twin system for short-term and long-term data analysis and closed-loop control.

[0189] (3) Data processing module: Receives the raw data table from the data transmission module, processes it and divides it into running batches, including: timestamp alignment, outlier removal, data filtering, and dividing the running window based on the equipment operating status and flow signal to obtain a data set organized by running batch.

[0190] And further includes:

[0191] The flow state calculation submodule is used to calculate the flow resistance state quantity, the baseline flow resistance state quantity, and the relative flow resistance state quantity. The relative flow resistance state quantity is used to characterize the changes in the rheological behavior of the material.

[0192] The multi-timescale trend analysis submodule is used to perform multi-timescale trend analysis on relative flow resistance state quantities, including the calculation of short-term and long-term relative flow resistance state quantities, in order to identify instantaneous fluctuations and overall trend changes in the extrusion process.

[0193] (4) Equipment status construction module: Receives the equipment operating parameters, relative flow resistance state quantity and multi-time scale trend analysis results output by the data processing module, integrates them, and constructs an equipment status table; the status table is used to characterize the dynamic operating status of the extrusion process and serves as the basic data for control.

[0194] (5) Process decision module: Receives the multi-timescale trend analysis results from the equipment status construction module and generates process suggestion information, including suggestions for water addition adjustment, feeding stability and material plasticization state adjustment, to guide operators to optimize the process.

[0195] (6) Blockage alarm analysis module: Receives the operation of the feeding motor and the feeding flow rate from the equipment status construction module, performs abnormal identification, outputs blockage warning or blockage alarm information when the threshold condition is met, and records the alarm time and duration.

[0196] (7) Front-end visualization module: Receives data output from the data processing module, equipment status construction module, and blockage alarm analysis module, and displays it through the digital twin platform interface. This includes equipment operating parameters, relative flow resistance state quantities, multi-timescale analysis results, and alarm information, realizing visualization and historical traceability analysis of production status. The module supports viewing historical trends at different time windows, enabling retrospective analysis of production status. Equipment operating parameters are obtained through the interface, including feed motor running (FeedMotorRunning, 1 / 0, running / stopping), feed flow (unit t / h), conditioner water flow (unit kg / h), conditioner steam flow (unit kg / h), and mechanical energy (unit kW.h / t). The short-term and long-term relative flow resistance state quantities calculated by the multi-timescale trend analysis submodule are displayed. , The system displays process recommendations in real time. The front end shows events such as blockage alarms and flow anomalies, including alarm type, start and end times, and number of alarms. Ideally, clicking on an alarm event will jump to the viscosity trend curve for the corresponding time point, helping equipment operators quickly locate problems. Alternatively, the front end can use chart components (such as line charts and sparklines) to display relative flow resistance state quantities and flow rate change trends. Historical trends and alarm events are overlaid, allowing operators to simultaneously observe process changes and potential anomalies. The front end polls the latest batch of equipment operating parameters through a unified interface. All data is bound to corresponding time windows, ensuring a one-to-one correspondence between data and actual production operation segments. Real-time refresh and interactive operations (selecting batches, jumping to specific time points, zooming to view trend curves) are supported.

[0197] (8) Prediction and optimization decision module (optional): used to build an HDMR prediction model based on the equipment status table, predict and evaluate candidate adjustment schemes, and determine the optimal adjustment scheme.

[0198] (9) Control execution module (optional): Combines proportional-integral controller and fuzzy neural network compensation to calculate control quantity and perform constraint processing to generate control instructions for equipment operation parameters and map them into physical quantity instructions that can be recognized by PLC. Adjust physical control components such as water valve and feed motor according to set value. The control execution result is collected again by the data acquisition module and fed back to the system to form closed-loop control.

[0199] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0200] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0201] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0202] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0203] The above embodiments are only used to illustrate the design concept and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, all equivalent changes or modifications made based on the principles and design ideas disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A method for constructing a digital twin of extrusion processing based on flow resistance state, characterized in that, Includes the following steps: S1: Collect equipment operating parameters to form continuous time series data, and obtain the set of operating batch time windows based on the judgment rules; S2: Collect equipment operating parameters in real time within the time window of the running batch to obtain the batch-level running dataset of the corresponding batch; S3: Calculate the flow resistance state variables The flow resistance state variables are obtained by outlier detection, filtering, and exponential smoothing. ; and then regarding the above or Monitoring continues until a steady-state condition is reached, under which the baseline flow resistance state quantity is obtained. ; S4: Based on the above , , Normalization is performed to construct the relative flow resistance state quantity. ; S5: Based on the above Construct trend states across multiple time scales; S6: Perform a comprehensive judgment on the multi-timescale trend status to generate process suggestion information and construct an equipment status table; S7 establishes an HDMR model and sets the target relative flow resistance state quantity, determines the optimal adjustment scheme, and then uses the equipment operating parameter group corresponding to the optimal adjustment scheme as the benchmark adjustment scheme for the current control input. It also performs feedback adjustment based on the deviation between the current relative flow resistance state quantity and the target relative flow resistance state quantity, and calculates the current control quantity. S8: The current control quantity is constrained to obtain the constrained control quantity and control instructions for the equipment operation parameters are generated. After receiving the control instructions, the PLC can adjust the physical control components to achieve real-time deviation correction.

2. The method according to claim 1, characterized in that, In step S1, the equipment operating parameters include the operating status of the feed motor, the venturi inlet pressure, the venturi outlet pressure, the feed flow rate, the water flow rate of the conditioner, the steam flow rate of the conditioner, and the mechanical energy. The specific determination rule is as follows: if equipment shutdown is detected, the downtime is accumulated. If the equipment restarts and the cumulative downtime is greater than or equal to the preset threshold, a new time window is determined to begin, and a unique identifier ID is generated. At the same time as the new batch begins, the previous window is marked as completed, and the cumulative output, flow rate, and operating status parameters of the current experimental window are initialized. When the equipment changes from the running state to the stopped state, or the feeding flow interruption lasts for more than the preset threshold, the current operating batch is determined to end.

3. The method according to claim 1, characterized in that, In step S3, The flow resistance state quantity is specifically the pressure difference between the Venturi inlet pressure and the Venturi outlet pressure. The stable operating condition specifically refers to: when or Within the preset time window The fluctuation range within is less than the threshold. The moment when this condition is first met is recorded as the candidate start time. The preset time window A continuous time interval is used to determine the stability of the flow resistance state quantity; then a stability determination process is performed, wherein the stability criterion for stability determination is: if from the candidate start time... Initially, the state quantity of flow resistance or Stability determination time The threshold is always met. The stability determination time If the time range is 20–60 s, the system is determined to have entered a stable operating condition, and the candidate start time is recorded. Backtracking confirmed as a stable start time If at the stability determination time If the stability criterion is not met at any given time, then the candidate start time is determined. If the system fails, a new candidate start time is searched, and the above stability determination process is repeated until the system enters a stable operating condition; wherein, the threshold... The range is 0.02–0.05 bar; The reference flow resistance state quantity Specifically: Regarding The number of sampling points is n Take the arithmetic mean, where, This is the baseline calculation time.

4. The method according to claim 1, characterized in that, In step S4, The normalization process specifically involves: if ,but For As molecules, with Fractions with a denominator of 1 are not 1; otherwise, they are written as 1. ;in, This is an invalid value.

5. The method according to claim 1, characterized in that, In step S5, The multi-timescale trend states include short-term relative flow resistance state quantities and long-term relative flow resistance state quantities. The short-term relative flow resistance state quantity Specifically: The difference; where, This indicates a short duration, ranging from 30 to 60 seconds. for The relative flow resistance state quantity before; The long-term relative flow resistance state quantity Specifically: The difference; where, Indicates a long duration, ranging from 240 to 360 seconds. for The relative flow resistance state quantity before.

6. The method according to claim 1, characterized in that, In step S6, The comprehensive judgment refers to: For the long term, if If the output process suggestion is: check the water supply to the equipment, and suggest reducing the water supply by 1-3%; if If the output process suggestion is: check the water addition adjustment, it is recommended to increase the water addition by 1-3%, and check the plasticization state of the material; Among them, the As the lower limit of the long-term trend change, the This represents the upper limit of the long-term trend change. In the short term, if Output process suggestion information: Check feeding stability; the aforementioned This is the lower limit of short-term trend change; if Output process suggestion information: Check feeding stability; the aforementioned This represents the upper limit of short-term trend changes; like and All are within the range of variation. Output process recommendation information: Maintain the current operating status and keep the original operating conditions. The equipment status table includes timestamps, relative flow resistance status values, multi-timescale trend status, equipment motor operating status, feed flow rate, water addition variation, steam flow rate, and unit energy consumption index.

7. The method according to claim 1, characterized in that, In step S7, The establishment of the HDMR model specifically involves: For the constant term of the HDMR model Individual device operating parameters right Impact Backup operating parameters and The interaction of Impact higher-order interaction pairs The sum of the effects of γ; where, The equipment operating parameter variables input to the HDMR model, i.e., the training samples X The target relative flow resistance state quantity is used to characterize the desired extrusion process state; The determination of the optimal adjustment scheme specifically involves: Calculate the predicted relative flow resistance state quantity corresponding to the k-th candidate adjustment scheme. : Set the candidate equipment operating parameter group corresponding to the k-th candidate adjustment scheme Substituting into the HDMR model, the candidate adjustment scheme is obtained by setting the adjustment range; Then compare the predicted relative flow resistance state quantities corresponding to the candidate adjustment schemes. Based on the deviation from the target relative flow resistance state quantity, the candidate adjustment scheme that makes the predicted relative flow resistance state quantity closest to the target state is selected as the optimal adjustment scheme.

8. The method according to claim 1, characterized in that, In step S8, The control instructions for obtaining constrained control quantities and generating equipment operating parameters by constraining the current control quantity are specifically as follows: The constrained control quantity is mapped into a physical quantity instruction recognizable by the PLC. This mapping process includes: determining the physical quantity form and unit based on the actuator type corresponding to each control variable, including feed flow rate (t / h), water addition change (t / h), and steam flow rate (kg / h); proportionally converting the control quantity according to the actuator's range to meet the PLC's setpoint range; discretizing the control quantity based on the actuator's minimum adjustment step size; performing necessary smoothing or speed limiting processing on the control signal based on the actuator's response characteristics; and encapsulating the processed control quantity into a PLC register or variable format and sending it to the PLC system for execution via an industrial communication protocol. The physical control components include a water inlet valve and a feed motor.

9. A digital twin system for a food extrusion processing procedure according to any one of claims 1-8, characterized in that, include: The data acquisition module is used to collect equipment operating parameters; The data transmission module is used to transmit the equipment operating parameters collected by the data acquisition module to form a raw data table; The data processing module processes the original data table and divides it into running batches to obtain a set of running batch time windows, and calculates the flow resistance state quantity, the baseline flow resistance state quantity, the relative flow resistance state quantity, and the short-term and long-term relative flow resistance state quantities. The equipment status construction module receives the equipment operating parameters, relative flow resistance state quantities, and multi-timescale trend analysis results output by the data processing module, integrates them, and constructs an equipment status table. The process decision module receives multi-timescale trend analysis results from the equipment status construction module and generates process recommendation information. The blockage alarm analysis module receives the operating status of the feeding motor and the feeding flow rate from the equipment status construction module, performs anomaly identification, and outputs blockage warning or blockage alarm information when the threshold conditions are met, and records the alarm time and duration. The front-end visualization module receives data from the data processing module, the equipment status construction module, and the congestion alarm analysis module, and displays it through the digital twin platform interface.

10. The system according to claim 9, characterized in that, The data acquisition module includes a feeding system, a conditioning system, a Venturi detection unit, an extrusion processing system, and a control and data processing system. The feeding system includes a feeding hopper, a feeding motor, and a flow sensor. The conditioning system includes a water addition unit, a steam heating unit, a temperature sensor, and a flow sensor. The Venturi detection unit includes a pressure sensor. The flow sensor, temperature sensor, and pressure sensor are specifically industrial sensors used to collect equipment operating parameters. The control and data processing system includes a PLC control unit and an edge computing unit used to mark the collected equipment operating parameters with a unified timestamp.

Images