Supercritical drying production line intelligent control method and system

Abstract

The application relates to the technical field of intelligent control, and provides a supercritical drying production line intelligent control method and system, which comprises the following steps: acquiring operation parameters of key equipment components in a production line, and determining performance change information of the key equipment components according to the operation parameters; acquiring measurement data of a plurality of associated sensors in the production line, and correcting the measurement data according to an associated relationship between physical quantities in the production line to obtain corrected measurement data; and adjusting a control strategy of the production line according to the performance change information and the corrected measurement data to generate a control instruction for compensating performance changes and measurement deviations. The application can effectively identify and quantify performance degradation of key equipment components and sensor measurement deviations in the supercritical drying production line, and dynamically adjust the control strategy accordingly, thereby overcoming the limitations of a traditional control system in the face of implicit equipment degradation and data drift, and significantly improving the accuracy, stability and product quality of production line operation.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology, and more specifically, to an intelligent control method and system for a supercritical drying production line. Background Technology

[0002] In the field of high-performance materials manufacturing, supercritical drying technology is widely used to prepare powder materials with specific microstructures due to its unique advantages. These production lines require extremely high precision in controlling temperature, pressure, and fluid flow to ensure product quality stability. However, in actual long-term operation, these production lines often face challenges that are not readily apparent, such as the subtle degradation of the performance of key equipment components and the drift in sensor measurement accuracy. These problems make it difficult for traditional control methods to maintain the accuracy and stability of the production process, thereby affecting product quality and energy efficiency.

[0003] For example, in a factory specializing in the production of high-performance composite materials, the core process involves using supercritical drying technology to prepare powder materials with specific microstructures. This production line consists of multiple independent supercritical drying reactors, each equipped with a heating jacket, high-pressure pump, heat exchanger, and carbon dioxide circulation system. The entire process requires extremely high precision in controlling temperature, pressure, and fluid flow; any minute deviation can cause fluctuations in key performance indicators such as particle size distribution, specific surface area, or porosity.

[0004] As the production line continued to operate, engineers began to notice some new and more subtle phenomena. For example, in the supercritical carbon dioxide circulation system, the heat exchanger's internal heat transfer surfaces began to show slight scaling or corrosion due to long-term operation, leading to a gradual decrease in heat transfer efficiency. This change was slow and gradual, and traditional control systems could not directly sense the physical state inside the heat exchanger; they could only determine whether to increase heating power based on feedback from the in-vessel temperature sensor. Due to the decrease in heat exchange efficiency, the control system had to instruct the heating system to output higher power or extend the heating time in order to maintain the in-vessel temperature at the set value.

[0005] This leads to mechanical wear and sealing degradation in critical actuators such as high-pressure pumps and valves under prolonged high-load operation. For example, increased internal leakage in high-pressure pumps results in actual carbon dioxide flow rates being lower than the control system's commanded values; or the response speed of regulating valves slows down, making precise fine-tuning of flow or pressure impossible. This degradation in mechanical performance further exacerbates the hysteresis and unresponsiveness of process parameters. When the control system receives signals that the temperature or pressure inside the reactor deviates from the target values, it issues more aggressive adjustment commands, attempting to compensate for this hysteresis by increasing power or opening degree. However, due to the deterioration in the performance of the actuators, these commands cannot be fully and effectively translated into the expected physical effects.

[0006] Worse still, this continuous compensatory operation makes the sensor data on the production line misleading. Some temperature, pressure, or flow sensors, under prolonged exposure to high pressure, high temperature, and fluid erosion, begin to exhibit slight drift or calibration deviations in measurement accuracy. When the control system bases its judgment on this drifting sensor data, it assumes the pressure inside the vessel has not yet reached the target, thus instructing the high-pressure pump to continue operating or delaying the depressurization process. However, the actual pressure inside the vessel may have already reached the target or even slightly exceeded it. This control based on inaccurate data not only leads to excessive energy consumption but may also cause powder materials to remain under undesirable pressure conditions for too long, affecting the integrity of their microstructure.

[0007] These hidden issues of equipment degradation and sensor drift have plunged the originally intelligent control system into a vicious cycle. It continuously issues seemingly reasonable compensation commands based on seemingly substandard parameters, but these commands often prove ineffective due to degraded performance of the actuators or are misdirected due to biased sensor data. As a result, the system, in order to maintain apparent parameter stability, is forced to operate continuously under overload, consuming energy far exceeding its design limits. This blind compensation not only causes enormous energy waste but also accelerates further aging of the equipment and shortens its lifespan.

[0008] Ultimately, this accumulation of persistent, latent deviations leads to further deterioration in product quality. For example, due to discrepancies between actual extraction time or pressure and set values, the residual solvent content in the powder material may be higher than expected, or the material structure may be damaged due to inaccurate pressure control during the depressurization process. These problems are only discovered during final product inspection, but by then it is difficult to trace back to the specific process step or equipment condition. Although the entire production line appears to be operating intelligently, it is actually working in a flawed state, with its operating efficiency, energy consumption, and product quality stability all falling far short of expectations. Furthermore, maintenance personnel struggle to accurately diagnose the root cause of the malfunctions and can only perform reactive repairs and replacements. Summary of the Invention

[0009] This application provides an intelligent control method and system for a supercritical drying production line to solve at least one of the above-mentioned technical problems.

[0010] To achieve the above objectives, this application adopts the following technical solution: In a first aspect, this application discloses an intelligent control method for a supercritical drying production line, used to address changes in the performance of equipment components and sensor measurement deviations during production line operation, comprising the following steps: Obtain the operating parameters of key equipment components in the production line, and determine the performance change information of key equipment components based on the operating parameters; Acquire measurement data from multiple related sensors in the production line, and correct the measurement data based on the measurement data and the correlation between physical quantities in the production line to obtain corrected measurement data; Based on performance change information and corrected measurement data, the production line control strategy is adjusted to generate control instructions to compensate for performance changes and measurement deviations.

[0011] Secondly, this application also discloses an intelligent control system for a supercritical drying production line, used to address changes in equipment component performance and sensor measurement deviations during production line operation. The system includes: The parameter acquisition module is used to acquire the operating parameters of key equipment components in the production line; The performance determination module is used to determine the performance change information of key equipment components based on operating parameters; The data acquisition module is used to acquire measurement data from multiple related sensors in the production line; The data correction module is used to correct the measurement data based on the relationship between the measurement data and the physical quantities in the production line, so as to obtain the corrected measurement data. The strategy adjustment module is used to adjust the production line's control strategy based on performance change information and corrected measurement data, in order to generate control instructions to compensate for performance changes and measurement deviations.

[0012] Compared with the prior art, this application has at least the following beneficial effects: This application acquires the operating parameters of key equipment components in the production line and determines their performance variation information. Simultaneously, it acquires and corrects measurement data from multiple related sensors, and finally adjusts the production line's control strategy based on this information to generate control commands that compensate for performance changes and measurement deviations. This method effectively solves problems in existing technologies such as decreased control accuracy, product quality fluctuations, and energy waste caused by the implicit degradation of equipment components and sensor measurement drift.

[0013] Specifically, by monitoring and quantifying the performance changes of key equipment components in real time, this application can identify potential equipment failures in advance, avoiding the blind compensation of traditional control systems when facing equipment degradation; at the same time, by intelligently correcting sensor measurement data, it overcomes the data misleading caused by sensor drift and ensures the accuracy of control decisions.

[0014] In summary, this application enables supercritical drying production lines to maintain high precision and stability of control during long-term operation, significantly improving product quality, reducing energy consumption, and extending equipment lifespan. This overcomes the limitations of existing technologies, such as difficulty in tracing the root cause of failures, passive maintenance, and low efficiency, and realizes intelligent and refined management of the production line. Attached Figure Description

[0015] Figure 1 A flowchart illustrating an intelligent control method for a supercritical drying production line provided in this application; Figure 2 This is a schematic diagram of the structure of an intelligent control system for a supercritical drying production line provided in this application. Detailed Implementation

[0016] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0017] like Figure 1 As shown in the embodiments of this application, an intelligent control method for a supercritical drying production line is proposed to address changes in equipment component performance and sensor measurement deviations during production line operation, including: Obtain the operating parameters of key equipment components in the production line, and determine the performance change information of key equipment components based on the operating parameters; Acquire measurement data from multiple related sensors in the production line, and correct the measurement data based on the measurement data and the correlation between physical quantities in the production line to obtain corrected measurement data; Based on performance change information and corrected measurement data, the production line control strategy is adjusted to generate control instructions to compensate for performance changes and measurement deviations.

[0018] The intelligent control method for supercritical drying production lines proposed in this application aims to effectively address the performance degradation of equipment components and sensor measurement deviations that occur during long-term operation of supercritical drying production lines through intelligent data analysis and strategy adjustment. Key equipment components refer to those that have a decisive impact on the stability, efficiency, and product quality of the supercritical drying process, such as high-pressure pumps, heat exchangers, and regulating valves. The operating status of these components directly affects whether the production line can operate stably according to preset process parameters. Operating parameters refer to real-time data reflecting the working status of these key equipment components, such as pump speed, current, and vibration; inlet and outlet temperatures and pressures of heat exchangers; and the opening degree of regulating valves. Performance change information refers to the deviation of equipment components from their design or baseline performance, quantified by analyzing operating parameters, such as the percentage decrease in heat transfer efficiency or the reduction in pump flow efficiency. Correlated sensors refer to sensors used in the production line to monitor different physical quantities (such as temperature, pressure, and flow rate), which have inherent physical correlations. For example, in a closed container, an increase in temperature is usually accompanied by an increase in pressure. The measurement data are the readings collected by these sensors in real time. The correlation between physical quantities refers to the physical laws or empirical models governing the relationship between different physical quantities in the production line, such as the ideal gas law and the law of conservation of energy. Corrected measurement data refers to measurement data that has been processed by algorithms to eliminate the influence of sensor bias or drift, and is closer to the true value. Control strategy refers to the preset control logic and parameters of the production line automation control system based on process requirements and real-time status. Control commands are the specific operational commands issued by the control system to actuators (such as heaters, pumps, and valves) to regulate the production line status.

[0019] Specifically, the control method of this application first focuses on the operating status of key equipment components in the production line. Various methods can be used to obtain the operating parameters of these key equipment components. For example, various sensors, such as vibration sensors, temperature sensors, pressure sensors, and current sensors, can be installed on the key equipment components to collect their operating data in real time. These sensors will continuously monitor the physical state of the equipment and transmit the data to the data acquisition system. Another method is to directly read the internal operating logs or status variables using the equipment's built-in control system interface. For example, modern high-pressure pumps typically record their operating time, cumulative flow, fault codes, etc., which can all be used as operating parameters. In addition, manual inspections can be used to periodically record the equipment's surface temperature, abnormal noises, etc., as supplementary operating parameters.

[0020] After obtaining these operating parameters, it is necessary to determine the performance changes of key equipment components based on these parameters. One approach is to compare the real-time operating parameters with the equipment's historical operating data or its baseline performance data at the time of manufacture. For example, if the outlet temperature of a heat exchanger remains consistently lower than the historical average under the same input heat, it may indicate a decrease in its heat transfer efficiency. Another approach is to establish a mathematical model of the equipment, input the real-time operating parameters into the model for simulation, and then compare the simulation results with the actual operating parameters. The deviation between the two can be used to infer performance changes. For example, an efficiency model for a high-pressure pump can be established. By inputting the pump's speed, outlet pressure, and flow rate, the theoretical efficiency can be calculated and then compared with the actual efficiency to quantify the pump's performance degradation.

[0021] This application also focuses on the measurement data of multiple related sensors in the production line. To acquire this data, a distributed sensor network can be deployed, connecting all critical sensors in the production line, such as those for temperature, pressure, flow rate, and liquid level, to a central data acquisition unit. These sensors send real-time readings to the central unit at a preset frequency (e.g., once per second or higher). For example, in a supercritical drying reactor, multiple temperature, pressure, and flow sensors are deployed simultaneously, and their data are acquired synchronously.

[0022] After acquiring the measurement data, it is necessary to correct it based on this data and the correlations between physical quantities in the production line to obtain corrected measurement data. One approach is to construct a soft sensor model using known correlations between physical quantities in the production line, such as the law of conservation of mass, the law of conservation of energy, or empirical formulas. When the reading of a sensor does not conform to the predictions of the physical model compared to the readings of other related sensors, the sensor can be considered to have a deviation, and the model can be used for correction. For example, in a closed supercritical drying vessel, there is a specific equation of state relationship between temperature, pressure, and density. If the temperature sensor reading increases, but the pressure sensor reading does not change accordingly, the equation of state, combined with the temperature reading, can be used to correct the pressure reading. Another approach is to use statistical methods, such as Kalman filtering or moving averages, to process the sensor data in real time to filter out noise and drift. For example, for a long-term stable temperature sensor, if its reading suddenly jumps sharply, and the readings of other related sensors do not change accordingly, the jump can be identified as an outlier, and historical data or data from adjacent time points can be used for smoothing correction.

[0023] Finally, based on the aforementioned performance change information and corrected measurement data, the production line's control strategy is adjusted to generate control commands to compensate for performance changes and measurement deviations. One implementation method is to input the performance change information and corrected measurement data into a pre-defined fuzzy logic controller or expert system. This system will determine the current production line status based on a pre-defined rule base and generate corresponding control commands. For example, if a 10% decrease in heat exchanger efficiency is detected, and the corrected reactor temperature is still below the target value, the system may instruct the heating system to increase heating power by 15% to compensate for the loss of heat transfer efficiency. Another approach is to employ a model-based predictive control (MPC) algorithm. This algorithm utilizes the production line's dynamic model, combined with performance change information and corrected measurement data, to predict the future state of the production line and optimize control commands to minimize deviations and meet process objectives. For example, if a decrease in the flow efficiency of the high-pressure pump is predicted, the MPC controller will adjust the pump's speed command in advance to ensure that the actual flow rate reaches the set value.

[0024] The intelligent control method for supercritical drying production lines proposed in this application significantly improves the stability and accuracy of production line operation by real-time sensing, quantification, and compensation for performance changes in key equipment components and sensor measurement deviations. Traditional control methods often rely solely on surface sensor readings for feedback control, failing to identify and address the implicit degradation of equipment components and sensor drift. This leads to blind compensation in the control system, resulting in energy waste and product quality fluctuations. This application, by introducing the determination of equipment component performance changes, enables the control system to gain a deeper understanding of the production line's health status, allowing for more targeted adjustments. Simultaneously, the correction of sensor measurement data ensures that the data upon which the control system bases its decisions is accurate and reliable, avoiding erroneous judgments due to data deviations. This dual compensation mechanism allows the production line to operate in an environment closer to ideal conditions, effectively preventing a vicious cycle caused by equipment degradation and sensor drift. For example, when the heat exchanger's heat transfer efficiency decreases, this application can identify this change and proactively adjust the heating power, rather than passively waiting for the temperature inside the vessel to deviate before making significant compensation. When the internal leakage of the high-pressure pump increases, this application can quantify the decrease in its flow efficiency and adjust the pump's commands in advance to ensure that the actual flow rate meets the target. This foresight and accuracy enable the production line to stably produce high-quality products with lower energy consumption and higher efficiency, significantly outperforming existing technologies that rely solely on surface parameter feedback.

[0025] However, in actual supercritical drying production lines, there are often complex interactions and dynamic relationships between key equipment components. Simply acquiring and analyzing the operating parameters of individual components independently may not be sufficient to comprehensively and accurately capture the performance degradation caused by the synergistic effects or mutual influences of multiple components, especially in the early stages. This degradation may manifest as subtle deviations in the overall response rather than significant anomalies in individual components. Failure to address these issues may result in untimely or inaccurate identification of performance change information, thereby affecting the effectiveness of control strategy adjustments.

[0026] In this regard, this application further proposes the following steps for obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters: Obtain the operating parameters of multiple key equipment components; Analyze the changing relationships between the operating parameters of multiple key equipment components in order to construct dynamic behavior patterns of these components; Compare dynamic behavior patterns with preset baseline behavior patterns to identify overall response deviations of multiple key equipment components; Decompose the overall response deviation to quantify the independent performance changes of multiple key equipment components.

[0027] Specifically, acquiring the operating parameters of multiple key equipment components means that the system simultaneously collects real-time operating data from multiple interrelated or collaborative key equipment components in the production line, such as temperature, pressure, flow rate, vibration frequency, current, and voltage. These parameters reflect the current operating status and performance of the components.

[0028] Analyzing the changing relationships among the operating parameters of multiple key equipment components to construct dynamic behavior patterns for these components can be understood as revealing the patterns of these parameters' changes over time and the mechanisms of their mutual influence by performing time series analysis, correlation analysis, or multivariate statistical analysis on the acquired operating parameters of multiple key equipment components. For example, techniques such as principal component analysis (PCA), independent component analysis (ICA), Kalman filtering, or neural network models can be used to establish a mathematical or statistical model that describes how these components collaboratively respond to external inputs or internal changes under normal operating conditions—that is, a dynamic behavior pattern. The aim is to capture the complex coupling relationships between components, rather than simply focusing on the independent performance of a single component.

[0029] Comparing dynamic behavior patterns with preset baseline behavior patterns to identify overall response deviations of multiple key equipment components involves comparing a dynamic behavior pattern established at the current moment or over a period of time with a baseline behavior pattern pre-established under healthy and stable production line operation. The baseline behavior pattern represents the ideal or expected collaborative working state of the components. By comparing, it can be found whether the overall response of multiple key equipment components in the current production line deviates from the normal range; this deviation is called overall response deviation. For example, the distance, similarity, or residual between the two patterns can be calculated to quantify this deviation. The purpose is to detect anomalies in the overall performance of the production line from a macroscopic perspective.

[0030] Decomposing the overall response deviation to quantify the independent performance changes of multiple key equipment components means that once the overall response deviation is identified, the system further uses diagnostic algorithms or models to trace this deviation back to specific, independent individual key equipment components. For example, contribution analysis, sensitivity analysis, or fault tree-based diagnostic methods can be used to analyze which component(s) performance degradation caused the overall deviation and quantify the degree of independent performance change of each component. The goal is to accurately pinpoint the source of the problem, providing a basis for subsequent refined control and maintenance.

[0031] This application's solution first acquires the operating parameters of multiple key equipment components and then analyzes the dynamic relationships between these parameters to construct a dynamic behavior pattern that reflects the complex coupling effects between components. It is precisely this modeling of the overall dynamic behavior that allows the system to transcend the independent analysis of single components and capture earlier, more subtle signs of performance degradation at the system level. Subsequently, by comparing the current dynamic behavior pattern with a preset baseline behavior pattern, the overall response deviation of the production line can be identified. This effectively solves the problem of inaccurate or untimely performance change identification that may result from neglecting the interactions between components. Finally, by decomposing the overall response deviation, the independent performance change information of each key equipment component can be precisely quantified, thereby avoiding the risk of attributing the overall problem to a faulty component and providing a solid foundation for subsequent adjustments to precise control strategies.

[0032] Through the above technical solution, this application can more comprehensively and accurately identify performance change information of key equipment components in supercritical drying production lines. Compared with methods that only focus on a single component, this solution, by considering the dynamic behavior patterns and interactions between multiple key equipment components, can detect potential performance degradation earlier and accurately locate specific components, thereby improving the sensitivity and accuracy of performance change identification. This provides a more reliable and refined basis for adjusting production line control strategies, effectively improving the stability and efficiency of production line operation and reducing the risk of unplanned downtime.

[0033] The following is a specific example to illustrate this.

[0034] In a supercritical drying production line, key equipment components include a high-pressure pump, a heater, and a separator. There is a close dynamic relationship between the operating parameters of these components (such as the outlet pressure of the high-pressure pump, the temperature of the heater, and the differential pressure of the separator).

[0035] First, the system will acquire operating parameters in real time, such as the outlet pressure of the high-pressure pump, the temperature of the heater, and the differential pressure of the separator.

[0036] Next, by analyzing time-series data of these parameters under normal operating conditions—for example, how the outlet pressure of the high-pressure pump and the differential pressure of the separator change in tandem as the heater temperature rises—a multivariate autoregressive (MAR) model is constructed as a dynamic behavior pattern for these components. This model can describe how these parameters interact and evolve over time under specific inputs.

[0037] Then, the parameters collected in real time during the current production line operation are input into the model, and its output dynamic behavior pattern is compared with the baseline MAR model established during the healthy operation period of the production line. For example, the differences in parameters between the two models or the statistical characteristics of the prediction residuals can be calculated to identify whether there is an overall response bias. If a significant difference is found between the current model and the baseline model, it indicates that the overall performance of the production line may have changed.

[0038] Finally, once the overall response deviation is identified, the system uses model-based sensitivity analysis to determine whether performance degradation in the high-pressure pump, heater, or separator contributes most to this deviation, and quantifies the independent performance changes of each component. For example, if a decrease in the efficiency of the high-pressure pump is found to cause a lower output pressure under the same input, thus affecting the overall dynamic response, the system will determine that there is a specific performance change in the high-pressure pump. In this way, even subtle overall deviations caused by interactions between components can be effectively identified and traced back to specific components, thereby achieving precise quantification of performance changes.

[0039] However, in actual operation, sensors may drift, or the physical model parameters used for correction may deviate from the actual situation due to environmental changes or long-term operation, resulting in deviations in the corrected measurement data, which in turn affects the accuracy and effectiveness of subsequent control strategies. If these problems are not addressed, the accuracy and robustness of production line control will be difficult to guarantee, potentially leading to product quality fluctuations or increased energy consumption. To address this, this application further proposes a more refined and adaptive measurement data correction method that improves the accuracy of measurement data by dynamically analyzing the sources of deviation and adjusting the physical model parameters.

[0040] In some embodiments, the steps of acquiring measurement data from multiple associated sensors in the production line, and correcting the measurement data based on the measurement data and the correlation between physical quantities in the production line to obtain corrected measurement data include: Collect real-time readings from multiple sensors; Predict expected values ​​using a physical model that includes adjustable parameters; Calculate the deviation between real-time readings from multiple sensors and predictions from the physical model; Analyze the dynamic patterns of the deviation to determine whether the deviation originates from sensor drift or deviation of physical model parameters; When the judgment deviation originates from the deviation of the physical model parameters, adjust the physical model parameters; The adjusted physical model, combined with real-time readings from multiple sensors, is used to correct the measurement data.

[0041] Specifically, acquiring real-time readings from multiple sensors refers to the system periodically obtaining the current measurement values ​​from various sensors deployed on the production line (such as temperature sensors, pressure sensors, flow sensors, etc.). These real-time readings form the basis for subsequent data correction.

[0042] Predicting expected values ​​using a physical model with adjustable parameters can be understood as constructing a mathematical model that describes the interactions between physical quantities in a production line, based on known relationships between these quantities. This physical model includes parameters that can be adjusted according to actual operating conditions, such as heat transfer coefficients and reaction rate constants. By inputting other relevant operating parameters of the current production line, the model can predict the expected measurement value of the sensor to be corrected under ideal conditions.

[0043] Calculating the deviation between real-time readings from multiple sensors and predictions from a physical model involves comparing the actual real-time sensor readings with the expected values ​​predicted by the physical model, quantifying the difference between the two. This deviation reflects the inconsistency between actual measurements and theoretical predictions.

[0044] Analyzing the dynamic patterns of deviations to determine whether they originate from sensor drift or deviations in physical model parameters aims to differentiate the root cause of measurement errors. For example, a slow, continuous, unidirectional trend in deviation may indicate sensor drift; a sudden increase in deviation under specific operating conditions or periodic fluctuations may suggest that certain parameters in the physical model are no longer accurate. This analysis can be achieved through statistical methods, signal processing techniques, or machine learning algorithms.

[0045] When the judgment deviation stems from the deviation of physical model parameters, the physical model parameters are adjusted to better reflect the actual operating status of the current production line. For example, optimization algorithms (such as least squares method, Kalman filtering, etc.) can be used to iteratively update the adjustable parameters in the physical model based on historical deviation data and real-time deviations, minimizing the deviation between the predicted and actual measured values.

[0046] Therefore, by using the adjusted physical model and combining real-time readings from multiple sensors, the measurement data can be corrected, with the aim of providing more accurate and reliable measurement data. Specifically, the more accurate expected values ​​provided by the adjusted physical model can be used, combined with real-time readings, and the original measurement data can be corrected through a fusion algorithm (such as weighted averaging, state observer, etc.), thereby eliminating or reducing errors caused by sensor drift and deviations in model parameters.

[0047] This application effectively addresses potential limitations in sensor data correction by introducing dynamic deviation analysis and an adaptive adjustment mechanism for physical model parameters. Specifically, firstly, a benchmark is constructed by acquiring real-time readings and predicting expected values ​​using a physical model containing adjustable parameters. Secondly, by calculating the deviation between real-time readings and predicted values ​​and further analyzing the dynamic patterns of this deviation, the system can intelligently distinguish whether the deviation is caused by sensor performance degradation (drift) or by inaccurate description of the current operating conditions by the physical model (parameter deviation). This intelligent judgment enables the system to take targeted measures: when model parameter deviation is identified, the system adaptively adjusts the physical model parameters, ensuring that the physical model always accurately reflects the actual physical relationships of the production line. Finally, the adjusted physical model is used to correct the measurement data, ensuring that the corrected data not only considers the relationships between physical quantities but also dynamically adapts to changes in the sensor and model itself, thus providing higher-precision input data.

[0048] Through the above technical solution, this application enables more accurate and robust correction of sensor measurement data in supercritical drying production lines. Compared to methods that rely solely on fixed physical correlations for correction, this application effectively avoids correction errors caused by sensor drift or model parameter deviations by dynamically identifying the source of deviation and adaptively adjusting the physical model parameters, significantly improving the accuracy and reliability of the measurement data. This provides a more solid and accurate data foundation for subsequent adjustments to production line control strategies, thereby enhancing the accuracy, stability, and adaptability of the entire production line control system, contributing to optimized product quality, reduced energy consumption, and extended equipment lifespan.

[0049] The following is a specific example to illustrate this.

[0050] In a supercritical drying production line, precise control of temperature and pressure within the drying chamber is required. The system deploys multiple temperature and pressure sensors and establishes a physical model describing the relationships between physical quantities such as temperature, pressure, and flow rate. This model includes adjustable parameters such as heat transfer coefficient and fluid resistance coefficient.

[0051] First, the system periodically collects real-time readings from these temperature and pressure sensors. Simultaneously, using current operating parameters such as flow rate and heating power, the system predicts the expected values ​​of temperature and pressure within the drying chamber through the aforementioned physical model.

[0052] Next, the system calculates the deviation between the real-time reading of each sensor and its corresponding expected value. For example, if the real-time reading of a temperature sensor is consistently higher than the expected value, and the deviation shows a slow upward trend, the system will analyze its dynamic pattern and determine that this may be due to drift of the temperature sensor. If there is a systematic deviation between the readings of all relevant sensors and the expected value, and the deviation suddenly increases after a specific process adjustment, the system may determine that this is because the heat transfer coefficient parameter in the physical model is no longer accurate.

[0053] When the system determines that the deviation mainly stems from the deviation of physical model parameters (such as the deviation of heat transfer coefficient), the system will initiate an online optimization algorithm to adjust the heat transfer coefficient in the physical model based on the current deviation data and historical data, so as to minimize the difference between its predicted value and the actual measured value.

[0054] Finally, this adjusted physical model, combined with real-time sensor readings, is used to correct the original measurement data. For example, data fusion techniques such as Kalman filtering can be used to fuse the adjusted model predictions with real-time readings to obtain more accurate temperature and pressure correction values. These corrected, high-precision data are then used to adjust control commands such as heating power and valve opening, thereby achieving precise control of the temperature and pressure within the drying chamber, ensuring the stability of the supercritical drying process and product quality.

[0055] When determining performance change information for key equipment components, the impact of production line operating scenarios on component performance may not be fully considered. This can lead to difficulties in accurately distinguishing between normal performance fluctuations and actual performance degradation based solely on operating parameters under different operating scenarios. Failure to address this issue may result in inaccurate judgment of performance change information, thereby affecting the effectiveness of control strategy adjustments. To address this, this application further proposes a process for obtaining operating parameters of key equipment components in the production line and determining performance change information based on these parameters. This process incorporates production line operating scenario information and baseline performance characteristics to more accurately assess component performance.

[0056] In some embodiments, the steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain production line operation context information; Based on production line operation context information, obtain the baseline performance characteristics of key equipment components under the production line operation context; Obtain the operating parameters of key equipment components under production line operating conditions; By comparing operating parameters with baseline performance characteristics, information on performance changes of key equipment components under production line operating conditions can be determined.

[0057] Production line operating context information can be understood as the external or internal environmental conditions that affect the performance of key equipment components, such as production load, ambient temperature, humidity, raw material batches, product types, and process parameter settings. Its purpose is to provide a specific and dynamic context for subsequent performance evaluation. In practical applications, production line operating context information can be obtained through various environmental sensors, process parameter monitoring systems, or production management systems deployed within the production line.

[0058] Benchmark performance characteristics refer to the expected performance of key equipment components under healthy or normal operating conditions in a specific production line operating scenario. These characteristics can be specific numerical ranges, curve models, or statistical distributions. For example, under specific loads and temperatures, the flow rate, head, and energy consumption of a pump, or the efficiency and vibration level of a compressor. Benchmark performance characteristics can be pre-established and stored through historical operating data analysis, equipment design specifications, performance curves provided by the manufacturer, experimental tests, or simulation models.

[0059] Comparing operating parameters with baseline performance characteristics involves comparing the currently acquired operating parameters of key equipment components with the baseline performance characteristics under the same or similar production line operating conditions. This comparison can employ various techniques, such as calculating the deviation between actual parameters and baseline characteristics, using statistical methods (e.g., mean and variance analysis), machine learning models (e.g., anomaly detection algorithms), or rule-based expert systems for judgment. The aim is to identify significant deviations between actual operating parameters and normal baseline performance, thereby indicating potential performance changes in the component.

[0060] This application introduces production line operating context information, enabling the assessment of performance changes in key equipment components to move beyond isolated assessments based on absolute values ​​or simple trends of operating parameters. Instead, it evaluates performance under specific operating conditions. Specifically, by acquiring the current production line operating context and retrieving or calculating the baseline performance characteristics of components within that context, a dynamic performance reference system consistent with actual operating conditions can be established. When actual operating parameters are compared with the baseline performance characteristics under this context, performance deviations from the normal range can be identified more accurately, effectively distinguishing between normal fluctuations caused by changes in the operating context and performance changes resulting from actual component degradation. This avoids misjudgments and ensures the accuracy of performance change information.

[0061] The above technical solution significantly improves the accuracy and reliability of determining performance changes in key equipment components. By considering the dynamic impact of production line operating conditions, this solution effectively avoids misjudgments caused by changes in operating conditions, making the identification of actual performance degradation of equipment components more accurate. This not only helps to promptly detect and handle potential equipment failures but also provides a more reliable basis for subsequent control strategy adjustments, thereby improving the operational stability and efficiency of the entire supercritical drying production line.

[0062] The following is a specific example to illustrate this.

[0063] Suppose a key component in a supercritical drying production line is a circulating pump used to transport CO2. The pump's performance (e.g., flow rate, head, energy consumption) is affected by operating conditions such as CO2 temperature, pressure, and the overall load of the production line. Basic solutions may only monitor the pump's flow rate or energy consumption; when these parameters change, it may be impossible to determine whether the change is due to pump performance degradation or simply normal fluctuations caused by changes in the production line load.

[0064] According to the scheme of this embodiment, the current production line operating situation information is first obtained. For example, the current CO2 temperature is 40°C, the pressure is 10MPa, and the production line load is 80%. Based on this situation information, the system will obtain the baseline performance characteristics that the circulating pump in a healthy state should have under this specific situation from a preset database or model. For example, under 40°C, 10MPa, and 80% load, the baseline flow rate of the pump should be XL / min, and the baseline energy consumption should be YkW.

[0065] The system then acquires the actual operating parameters of the circulating pump, such as the actual flow rate (X' L / min) and the actual energy consumption (Y' kW). By comparing the actual operating parameters (X', Y') with the baseline performance characteristics (X, Y), if the actual flow rate is significantly lower than the baseline flow rate, or the actual energy consumption is significantly higher than the baseline energy consumption, and this deviation exceeds the preset normal fluctuation range, then the system can accurately determine that there is a performance change in the circulating pump. For example, if X' is much smaller than X, it may indicate a decrease in pump efficiency or the presence of leakage; if Y' is much larger than Y, it may indicate increased pump wear or blockage.

[0066] In this way, the solution can accurately identify the actual performance changes of key equipment components under specific operating conditions, avoiding misjudgments caused by changes in conditions, and providing more accurate input for subsequent control strategy adjustments.

[0067] When acquiring operating parameters and determining performance changes of key equipment components in a production line, the combined impact of external environmental factors and upstream processes on these parameters may not be fully considered. This can lead to inaccurate performance change information, affecting the effectiveness of subsequent control strategy adjustments. Failure to address this issue could result in misjudgments or omissions of equipment component performance degradation, leading to production line instability, product quality fluctuations, or even equipment failures. To address this, this application proposes a more comprehensive method that effectively separates the influence of interfering factors through integrated analysis of multi-source operating parameters and environmental information, thereby more accurately determining the performance changes of key equipment components.

[0068] In some embodiments, the steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Acquire multi-source operating parameters and environmental information, including operating parameters of key equipment components, external environmental parameters of the production line, and upstream process parameters; Analyze the correlation between multi-source operating parameters and environmental information to extract the impact patterns of interference sources; Based on the influence mode of the interference source, the multi-source operating parameters are processed to separate the parameter changes caused by the performance degradation of key equipment components; Quantify parameter changes to determine performance variations in key equipment components.

[0069] Multi-source operating parameters and environmental information refer to comprehensive data obtained from multiple sources that reflect the production line's operating status and external conditions. Specifically, operating parameters of key equipment components are indicators that directly reflect the working status of the equipment components, such as temperature, pressure, current, vibration, and speed; external environmental parameters of the production line refer to the external environmental conditions in which the production line operates, such as ambient temperature, humidity, and atmospheric pressure; upstream process parameters refer to the operating indicators of one or more previous process stages that affect the input conditions of the current production line segment, such as feed flow rate, material concentration, and pretreatment temperature. This multi-source information is collected for comprehensive analysis.

[0070] Analyzing the correlation between multi-source operating parameters and environmental information to extract the influence patterns of interference sources refers to identifying the interaction relationships between different parameters through statistical methods, machine learning algorithms, or analysis based on physical models. For example, regression models, neural network models, or causal graphs can be established to reveal how external environmental parameters or upstream process parameters affect the operating parameters of key equipment components. This allows for the identification of potential external factors as interference sources, as well as their specific ways and degrees of influence on the operating parameters of key equipment components, thus forming the influence patterns of interference sources.

[0071] Based on the influence patterns of interference sources, multi-source operating parameters are processed to separate parameter changes caused by performance degradation of key equipment components. Specifically, this processing may include filtering, compensating, or modeling the original operating parameters to remove the influence of identified interference sources. For example, if it is known that an increase in ambient temperature will cause a generally higher reading for a certain pressure sensor, the pressure reading can be corrected according to the influence pattern of ambient temperature to obtain a pure pressure value unaffected by ambient temperature. The aim is to distinguish parameter fluctuations caused by external interference from parameter changes caused by the performance degradation of the equipment components themselves.

[0072] Ultimately, parameter changes are quantified to determine performance variations in critical equipment components. After isolating the effects of confounding factors, the remaining parameter changes are considered to more accurately reflect the performance degradation of critical equipment components. These parameter changes can be quantified into specific performance indicators, such as percentage efficiency reduction, wear index, or remaining life prediction, thus forming performance variation information for critical equipment components.

[0073] This application comprehensively captures various factors affecting the operating status of equipment components by acquiring multi-source operating parameters and environmental information, including operating parameters of key equipment components, external environmental parameters of the production line, and upstream process parameters. By analyzing the correlation between these multi-source information, the influence patterns of interference sources not caused by component degradation, such as external environment or upstream processes, can be identified and extracted. Based on these influence patterns, the original operating parameters are refined, effectively separating parameter changes caused by performance degradation of key equipment components from parameter changes caused by external interference. Therefore, the quantified parameter changes more accurately reflect the actual performance degradation of key equipment components, avoiding misjudgments caused by external interference and providing a more reliable basis for adjusting subsequent control strategies.

[0074] Through the above technical solution, this application overcomes the limitations of determining the performance changes of key equipment components, which is easily affected by external environmental and upstream process interference factors. By introducing multi-source operating parameters and environmental information, and employing advanced analysis methods to separate the influence of interference sources, the determined performance change information becomes more accurate and reliable. This not only improves the accuracy of assessing the actual health status of equipment components but also provides a solid foundation for the refined adjustment of production line control strategies, thereby effectively improving the overall operational stability and efficiency of the supercritical drying production line and reducing maintenance costs and downtime risks caused by misjudgments.

[0075] Here, we will take the high-pressure pump in the supercritical drying production line as an example for illustration.

[0076] The operating parameters of this high-pressure pump include its speed, current, outlet pressure, and vibration data. Simultaneously, external environmental parameters of the production line are collected, such as workshop ambient temperature and humidity, as well as upstream process parameters, such as the moisture content and flow rate of the material entering the drying tower. When the outlet pressure of the high-pressure pump fluctuates, the basic approach might directly attribute it to a decline in pump performance. However, the approach in this application first analyzes the correlation between these multi-source operating parameters and environmental information. For example, it was found that when the workshop ambient temperature rises, the cooling efficiency of the high-pressure pump decreases, leading to an increase in its internal temperature, which in turn affects its outlet pressure reading. By establishing an interference model between ambient temperature and outlet pressure, the influence pattern of ambient temperature on outlet pressure can be extracted. Subsequently, based on this influence pattern, the outlet pressure data of the high-pressure pump is processed to remove pressure fluctuations caused by changes in ambient temperature. Finally, the remaining outlet pressure changes caused by the wear or aging of the high-pressure pump itself are quantified, thereby more accurately determining the performance change information of the high-pressure pump, such as the actual degree of decline in its pumping efficiency. Based on this precise information on performance changes, the control system can generate more reasonable control commands, such as fine-tuning the pump speed or arranging preventative maintenance, to compensate for performance degradation and maintain stable production line operation.

[0077] However, in practical applications, timely and accurate identification of early performance degradation in critical equipment components, and quantification of this degradation into information that can be used to adjust control strategies, is a key challenge in ensuring stable production line operation. Relying solely on simple parameter threshold judgments may fail to capture subtle performance change trends, leading to delayed warnings or false alarms. To address this, this application proposes a more refined method that uses sequence analysis and feature comparison of operating parameters to effectively identify and quantify early performance changes in critical equipment components.

[0078] In some embodiments, the steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain the operating parameters of key equipment components; The acquired operating parameters are preprocessed to obtain a smooth sequence of operating parameters; Calculate the short-term and long-term rates of change of the operating parameter sequence based on the smoothed operating parameter sequence; Establish a benchmark characteristic range that reflects the health status of key equipment components; The calculated short-term and long-term rates of change of operating parameters are compared with the baseline characteristic range to trigger an early warning signal; Quantify early warning signals to identify early performance changes in critical equipment components.

[0079] Obtaining the operating parameters of key equipment components refers to collecting various indicators related to the operating status of key equipment components in real time or periodically through sensors, data acquisition systems, etc., such as temperature, pressure, vibration, current, voltage, and rotational speed. These parameters are the basic data for evaluating component performance.

[0080] Preprocessing the acquired operating parameters to obtain a smooth operating parameter sequence involves performing data cleaning, noise reduction, and filtering on the raw collected operating parameters to eliminate measurement errors, transient interference, or outliers, thereby obtaining a more stable and representative data sequence. For example, moving averages, exponential smoothing, or Kalman filtering can be used to process the data to reduce random fluctuations and highlight data trends.

[0081] Calculating the short-term and long-term rates of change of operating parameters, based on a smoothed sequence of operating parameters, involves using mathematical methods such as differencing, regression analysis, or time series models to calculate the rate of change of parameters within a shorter time window and their trend within a longer time window. Short-term rates of change can reflect instantaneous fluctuations or sudden anomalies in component performance, while long-term rates of change reveal the slow degradation or aging trend of component performance.

[0082] Establishing a baseline characteristic range reflecting the health status of critical equipment components refers to setting a normal range or healthy interval for the short-term and long-term rates of change of key operating parameters, based on historical operating data, design specifications, expert experience, or failure mode analysis. This interval defines the expected behavior pattern of the component under normal operating conditions.

[0083] The method of comparing the calculated short-term and long-term rates of change of operating parameters with the baseline characteristic range to trigger an early warning signal involves comparing the real-time calculated short-term and long-term rates of change with the preset baseline characteristic range. When any rate of change exceeds its corresponding baseline characteristic range, it indicates that the component performance may be abnormal or degraded, and the system will trigger an early warning signal.

[0084] Quantifying early warning signals to identify early performance changes in critical equipment components involves further analyzing and evaluating the triggered warning signals, transforming them into specific and actionable performance change information. For example, warning signals can be quantified into different performance degradation levels or risk indices based on factors such as the degree to which they exceed the baseline range, their duration, and the absolute value of the rate of change, thereby providing a refined basis for subsequent adjustments to control strategies.

[0085] This application effectively addresses the limitations in identifying early performance degradation of critical equipment components by introducing preprocessing of operating parameters, calculation of short-term and long-term change rates, and comparison with benchmark characteristic intervals. Specifically, preprocessing the raw operating parameters eliminates noise interference, ensuring the accuracy of subsequent analysis. Calculating short-term change rates allows the system to promptly capture instantaneous anomalies or rapid changes in component performance, while long-term change rates help identify slow and continuous degradation trends. Comparing these change rates with pre-established benchmark characteristic intervals objectively determines whether the current health status of the component deviates from the normal range, thus triggering early warnings in the early stages of performance degradation. Finally, quantifying the warning signal transforms vague warnings into specific performance change information, providing data support for precise adjustments to production line control strategies and preventing production line efficiency decline or malfunctions caused by performance degradation.

[0086] Through the above technical solution, this application enables the accurate identification and quantification of early performance changes in key equipment components in supercritical drying production lines. Compared to methods relying solely on fixed thresholds or single parameters, this solution significantly improves the sensitivity and timeliness of early warning for potential component failures or performance degradation trends by comprehensively analyzing the short-term and long-term trends of operating parameters and comparing them with dynamic benchmark characteristic ranges. This allows the production line control system to obtain early performance change information before significant component performance deterioration, thereby enabling targeted adjustments to control strategies, effectively compensating for performance changes, and avoiding reduced production line efficiency, product quality fluctuations, or even unexpected shutdowns caused by slow component performance degradation, significantly improving the operational stability and reliability of the production line.

[0087] The following is a specific example to illustrate this.

[0088] Suppose that a key component in a supercritical drying production line is a high-pressure pump, whose operating parameters include outlet pressure, motor current, and vibration frequency.

[0089] First, the system continuously acquires operating parameters such as the outlet pressure of the high-pressure pump, motor current, and vibration frequency.

[0090] Next, these acquired operating parameters are preprocessed, for example, by using moving average filtering to eliminate sensor noise and instantaneous fluctuations, resulting in a smooth sequence of operating parameters.

[0091] Then, based on these smoothed sequences, their short-term rates of change (e.g., changes per minute) and long-term rates of change (e.g., average trends per hour or day) are calculated. For example, the instantaneous rate of pressure drop and the daily average pressure drop trend at the outlet pressure can be calculated.

[0092] Meanwhile, based on historical health operation data of the high-pressure pump and performance indicators provided by the manufacturer, a benchmark characteristic range reflecting the health status of the high-pressure pump is established. For example, under normal circumstances, the short-term rate of change of outlet pressure should be within ±0.5 MPa / min, and the long-term rate of change should be within ±0.1 MPa / day.

[0093] The system will then compare the real-time calculated short-term and long-term rates of change with these benchmark characteristic ranges. If the long-term rate of change of the outlet pressure is found to continuously exceed -0.1 MPa / day (i.e., the pressure continues to decrease slowly), or the short-term rate of change of the motor current suddenly increases, the system will trigger an early warning signal.

[0094] Finally, the triggered early warning signals are quantified. For example, if the long-term rate of change in outlet pressure consistently exceeds the baseline range, the system can quantify it as performance change information in the early stages of high-pressure pump seal wear and provide a corresponding degradation level. This quantified early performance change information is then used to adjust the production line's control strategy. For instance, the operating frequency of the high-pressure pump can be fine-tuned or maintenance plans can be scheduled in advance to compensate for pressure losses caused by seal wear, thereby ensuring stable production line operation.

[0095] However, in actual production line operation, the performance degradation of equipment components is often a gradual process, and its performance changes are not singular or linear. If the degradation stage of a component is not accurately identified and evaluated using corresponding patterns, the determination of performance change information may be inaccurate, thus affecting the timeliness and effectiveness of control strategy adjustments. To address this, this application proposes a more refined method for determining the performance change information of key equipment components. This method identifies the degradation stage of a component and combines it with specific performance degradation patterns to more accurately quantify performance changes.

[0096] In some embodiments, the steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain the operating parameters of key equipment components; The acquired operating parameters are preprocessed to obtain characteristic data reflecting the state of the components; Analyze feature data to identify the current degradation stage of the operating parameters of critical equipment components; Based on the identified degradation stage, retrieve the performance degradation mode specific to that degradation stage; Based on the performance degradation pattern and combined with feature data, the performance change information of key equipment components is determined.

[0097] Specifically, acquiring the operating parameters of key equipment components refers to collecting various data related to the operating status of key equipment components in real time or periodically through sensors, data acquisition systems, etc., such as temperature, pressure, vibration, current, voltage, speed, flow rate, etc. These parameters directly reflect the immediate working status of the components.

[0098] Preprocessing the acquired operating parameters yields feature data reflecting the component's state. This can be understood as performing data cleaning, noise reduction, normalization, and feature extraction on the raw operating parameters. For example, moving averages and Kalman filters can be used to remove noise, or principal component analysis and wavelet transform can be used to extract key features that effectively characterize the component's health status from multidimensional operating parameters. The aim is to eliminate redundancy and interference in the raw data and highlight the core information related to component performance degradation.

[0099] Analyzing feature data to identify the current degradation stage of critical equipment components involves using machine learning models (such as support vector machines, neural networks, and decision trees) or statistical analysis methods (such as threshold judgment and trend analysis) based on preprocessed feature data to classify the component's operating status into different degradation stages, such as normal operation, early degradation, intermediate degradation, and severe degradation. The aim is to implement tiered management of component health status, providing a basis for subsequent accurate assessments.

[0100] Based on the identified degradation stage, the specific performance degradation patterns for that stage are retrieved. This refers to pre-establishing or learning corresponding performance degradation models or pattern libraries for different degradation stages. For example, in the early degradation stage, a component's performance may exhibit slight efficiency decline or localized wear; while in the severe degradation stage, it may exhibit significant performance degradation or increased failure risk. These patterns can be mathematical models, empirical curves, or statistical distributions based on historical data. The aim is to provide a targeted assessment framework for performance changes at different degradation stages.

[0101] Therefore, determining the performance change information of key equipment components based on performance degradation patterns and characteristic data involves comparing and calculating the current characteristic data of the component with the performance degradation pattern of the corresponding degradation stage, thereby quantifying the degree of performance deviation. For example, it can calculate the deviation between the current performance and the baseline performance of the degradation stage, or predict the remaining lifespan of the component under the current degradation trend. The purpose is to provide accurate and dynamic performance change information to support subsequent adjustments to control strategies.

[0102] This application first acquires the operating parameters of key equipment components and preprocesses them to obtain characteristic data reflecting the component's state, effectively filtering out noise and irrelevant information and focusing on the core health indicators of the components. It is precisely this in-depth analysis of the characteristic data that allows for the identification of the specific degradation stage of the key equipment components. This enables the evaluation of component performance to move beyond a single dimension and instead be stratified based on the actual health status of the components. Based on this, the application retrieves the performance degradation patterns specific to each identified degradation stage, ensuring the relevance and accuracy of the performance evaluation model and avoiding errors that may arise from a one-size-fits-all approach. Finally, based on this staged performance degradation pattern and the current characteristic data, the performance change information of key equipment components can be more accurately determined, providing a more reliable and refined input for subsequent control strategy adjustments.

[0103] Through the above technical solution, this application overcomes the limitations in handling the progressive performance degradation of equipment components. By introducing feature data extraction and degradation stage identification, the perception of component performance changes becomes more sensitive and accurate, enabling timely detection and quantification of early degradation signs. Furthermore, by retrieving specific performance degradation modes for different degradation stages, the accuracy and reliability of performance change information determination are significantly improved, avoiding misjudgments or delays caused by model mismatch. Consequently, the generated control commands can more accurately compensate for component performance changes, thereby effectively improving the operational stability, efficiency, and safety of the supercritical drying production line, extending equipment lifespan, and reducing the risk of unplanned downtime.

[0104] The following is a specific example to illustrate this.

[0105] In a supercritical drying line, a key component is a high-pressure pump. Its operating parameters include pump outlet pressure, flow rate, motor current, vibration frequency, and bearing temperature. First, the system continuously acquires these operating parameters. Next, these raw parameters are preprocessed; for example, Fourier transform is performed on the vibration signal to extract vibration amplitudes at specific frequencies as feature data, and moving averages are applied to the temperature and current data to smooth noise. Then, a pre-trained classification model (such as a support vector machine based on historical fault data) is used to analyze this feature data to identify whether the high-pressure pump is currently in a normal wear stage, a slight leakage stage, or an early stage of bearing wear. Once the early stage of bearing wear is identified, the system retrieves the performance degradation pattern specific to this stage. This pattern might be a mathematical function describing the accelerated rise in bearing temperature over operating time, or a trend model based on changes in the vibration spectrum. Finally, based on this specific degradation pattern, combined with the current bearing temperature and vibration feature data, the system accurately calculates the degree of performance degradation of the high-pressure pump, for example, quantifying its efficiency loss percentage or predicting its remaining healthy life. This performance change information is then transmitted to the control strategy adjustment module to adjust the pump's operating frequency or maintenance schedule in a timely manner.

[0106] like Figure 2 As shown in the embodiments of this application, an intelligent control system for a supercritical drying production line is also disclosed, which is used to cope with changes in the performance of equipment components and sensor measurement deviations during production line operation. The system includes: Parameter acquisition module 1 is used to acquire the operating parameters of key equipment components in the production line; Performance determination module 2 is used to determine the performance change information of key equipment components based on operating parameters; Data acquisition module 3 is used to acquire measurement data from multiple related sensors in the production line; The data correction module 4 is used to correct the measurement data based on the relationship between the measurement data and the physical quantities in the production line, so as to obtain the corrected measurement data. Strategy adjustment module 5 is used to adjust the control strategy of the production line based on performance change information and corrected measurement data, so as to generate control instructions to compensate for performance changes and measurement deviations.

[0107] The intelligent control system for a supercritical drying production line provided in this application aims to effectively address the performance degradation of equipment components and sensor measurement deviations that occur during long-term operation of a supercritical drying production line through modular design. The parameter acquisition module is responsible for collecting real-time operating data of key equipment components in the production line, providing a basis for subsequent analysis. The performance determination module quantifies the deviation of equipment components from their design or benchmark performance based on these operating parameters, thereby gaining a deeper understanding of the production line's health status. Simultaneously, the data acquisition module collects real-time measurement data from various related sensors in the production line, while the data correction module uses the inherent correlation between physical quantities to correct these measurement data, ensuring that the data upon which the control system's decisions are based is accurate and reliable. Finally, the strategy adjustment module comprehensively utilizes performance change information and corrected measurement data to intelligently adjust the production line's control strategy, generating precise control commands to compensate for equipment performance changes and measurement deviations, thereby maintaining the stability and accuracy of the production line's operation and avoiding the vicious cycle caused by equipment degradation and sensor drift in traditional control methods.

[0108] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for intelligent control of a supercritical drying production line, characterized in that, Used to address changes in equipment component performance and sensor measurement deviations during production line operation, including: Obtain the operating parameters of key equipment components in the production line, and determine the performance change information of key equipment components based on the operating parameters; Acquire measurement data from multiple related sensors in the production line, and correct the measurement data based on the measurement data and the correlation between physical quantities in the production line to obtain corrected measurement data; Based on performance change information and corrected measurement data, the production line control strategy is adjusted to generate control instructions to compensate for performance changes and measurement deviations.

2. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain the operating parameters of multiple key equipment components; Analyze the changing relationships between the operating parameters of multiple key equipment components in order to construct dynamic behavior patterns of these components; Compare dynamic behavior patterns with preset baseline behavior patterns to identify overall response deviations of multiple key equipment components; Decompose the overall response deviation to quantify the independent performance changes of multiple key equipment components.

3. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of acquiring measurement data from multiple associated sensors in the production line, and correcting the measurement data based on the measurement data and the correlation between physical quantities in the production line to obtain corrected measurement data include: Collect real-time readings from multiple sensors; Predict expected values ​​using a physical model that includes adjustable parameters; Calculate the deviation between real-time readings from multiple sensors and predictions from the physical model; Analyze the dynamic patterns of the deviation to determine whether the deviation originates from sensor drift or deviation of physical model parameters; When the judgment deviation originates from the deviation of the physical model parameters, adjust the physical model parameters; The adjusted physical model, combined with real-time readings from multiple sensors, is used to correct the measurement data.

4. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain production line operation context information; Based on production line operation context information, obtain the baseline performance characteristics of key equipment components under the production line operation context; Obtain the operating parameters of key equipment components under production line operating conditions; By comparing operating parameters with baseline performance characteristics, information on performance changes of key equipment components under production line operating conditions can be determined.

5. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Acquire multi-source operating parameters and environmental information; Analyze the correlation between multi-source operating parameters and environmental information to extract the impact patterns of interference sources; Based on the influence mode of the interference source, the multi-source operating parameters are processed to separate the parameter changes caused by the performance degradation of key equipment components; Quantify parameter changes to determine performance variations in key equipment components.

6. The intelligent control method for a supercritical drying production line according to claim 5, characterized in that, The acquisition of multi-source operating parameters and environmental information includes operating parameters of key equipment components, external environmental parameters of the production line, and upstream process parameters.

7. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain the operating parameters of key equipment components; The acquired operating parameters are preprocessed to obtain a smooth sequence of operating parameters; Calculate the short-term and long-term rates of change of the operating parameter sequence based on the smoothed operating parameter sequence; Establish a benchmark characteristic range that reflects the health status of key equipment components; The calculated short-term and long-term rates of change of operating parameters are compared with the baseline characteristic range to trigger an early warning signal; Quantify early warning signals to identify early performance changes in critical equipment components.

8. The intelligent control method for a supercritical drying production line according to claim 1, characterized in that, The steps of obtaining the operating parameters of key equipment components in the production line and determining the performance change information of the key equipment components based on the operating parameters include: Obtain the operating parameters of key equipment components; The acquired operating parameters are preprocessed to obtain characteristic data reflecting the state of the components; Analyze feature data to identify the current degradation stage of the operating parameters of critical equipment components; Based on the identified degradation stage, retrieve the performance degradation mode specific to that degradation stage; Based on the performance degradation pattern and combined with feature data, the performance change information of key equipment components is determined.

9. The intelligent control method for a supercritical drying production line according to claim 8, characterized in that, The acquired operating parameters are preprocessed to obtain feature data reflecting the component status. This includes data cleaning, noise reduction, normalization, and feature extraction of the acquired operating parameters to obtain feature data reflecting the component status.

10. An intelligent control system for a supercritical drying production line, characterized in that, Used to address changes in equipment component performance and sensor measurement deviations during production line operation, including: The parameter acquisition module is used to acquire the operating parameters of key equipment components in the production line; The performance determination module is used to determine the performance change information of key equipment components based on operating parameters; The data acquisition module is used to acquire measurement data from multiple related sensors in the production line; The data correction module is used to correct the measurement data based on the relationship between the measurement data and the physical quantities in the production line, so as to obtain the corrected measurement data. The strategy adjustment module is used to adjust the production line's control strategy based on performance change information and corrected measurement data, in order to generate control instructions to compensate for performance changes and measurement deviations.

Images