A method and system for controlling the flow rate of cleaning solution for wine fermentation valve array

Abstract

This invention relates to the technical field of cleaning fluid flow control, specifically to a method and system for controlling the flow of cleaning fluid in a wine fermentation valve array. The method includes the following steps: establishing a correspondence between cleaning pump operating parameters, regulating valve opening parameters, and cleaning fluid flow rate; calculating a reference value for the cleaning fluid flow rate; evaluating the consistency of the cleaning fluid flow rate measurement signal, the reference value, and pipeline pressure parameters; determining that the cleaning fluid flow rate measurement signal is unreliable when the deviation between the measurement signal and the reference value exceeds a preset flow deviation threshold, and the regulating valve opening parameter or pipeline pressure parameter exceeds a preset threshold range; switching the control mode when the measurement signal is deemed unreliable, and controlling the cleaning fluid flow rate based on the reference value in the new control mode; and issuing a maintenance alarm. This method helps improve the accuracy of detection.

Description

Technical Field

[0001] This invention relates to the technical field of cleaning fluid flow control, and specifically to a method and system for controlling the flow of cleaning fluid for a wine fermentation valve array. Background Technology

[0002] In the modern beer brewing industry, fermentation tanks and their connected valve arrays require thorough cleaning after fermentation to ensure product quality and hygienic production conditions. Flow control of the cleaning solution is crucial for efficient and resource-saving cleaning. Typically, electromagnetic flow meters are used to measure the flow rate and control it based on pre-set parameters. However, for cost control and sustainability reasons, the cleaning solution is often recycled. This recycling leads to cumulative changes in the physical properties of the cleaning solution, such as the dissolution and suspension of organic matter, which in turn affects the measurement accuracy of the electromagnetic flow meter. Summary of the Invention

[0003] The purpose of this invention is to address the aforementioned shortcomings by proposing a method and system for controlling the flow rate of cleaning liquid in wine fermentation valve arrays.

[0004] The present invention adopts the following technical solution:

[0005] A method for controlling the flow rate of cleaning fluid for wine fermentation valve arrays, the method comprising the following steps:

[0006] Collect cleaning fluid flow measurement signals, cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters;

[0007] Establish the correspondence between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid; and calculate the reference value of the cleaning fluid flow rate based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve.

[0008] The consistency of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameter is evaluated. When the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds the preset flow deviation threshold, and the regulating valve opening parameter or the pipeline pressure parameter exceeds the preset threshold range, the cleaning fluid flow measurement signal is determined to be unreliable.

[0009] When the cleaning fluid flow measurement signal is determined to be unreliable, the control mode is switched, and the cleaning fluid flow is controlled according to the cleaning fluid flow reference value in the switched control mode.

[0010] A maintenance alert has been issued.

[0011] This technical solution can effectively identify unreliable situations in the cleaning fluid flow measurement signal and switch the control mode in a timely manner. It controls the flow based on the calculated reference value of the cleaning fluid flow, thereby avoiding excessive scouring and wear of the valve array system caused by inaccurate flow meter measurement. It solves the problem of difficult detection of hidden equipment damage in existing technologies and can issue maintenance alarms in a timely manner to remind users to carry out maintenance.

[0012] Furthermore, the steps for conducting a conformity assessment include:

[0013] Obtain the opening command signal of the control valve;

[0014] The opening command signal of the control valve is subjected to high-pass filtering to obtain the rapidly changing components of the command signal;

[0015] Calculate the root mean square value of the rapidly changing component of the command signal to obtain the quantitative index of the rapidly changing component of the command signal.

[0016] Based on the quantification index of the rapidly changing components of the command signal, the proportional gain and derivative gain of the controller used for consistency evaluation are dynamically adjusted.

[0017] When the quantization index of the rapidly changing components of the command signal exceeds the preset upper limit threshold, the proportional gain and derivative gain of the controller are reduced.

[0018] When the quantization index of the rapidly changing component of the command signal remains below the preset upper limit threshold, the proportional gain and derivative gain of the controller are restored.

[0019] Based on the adjusted controller, a consistency assessment is performed on the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameters.

[0020] Furthermore, the steps for calculating the reference value of the cleaning fluid flow rate include:

[0021] Periodically perform a performance self-calibration procedure on the cleaning pump and control valve;

[0022] At the preset operating point, measure the actual output pressure and actual input power of the cleaning pump;

[0023] The theoretical flow rate value is obtained based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve;

[0024] Determine whether there is a preset allowable deviation threshold between the actual output pressure or actual input power of the cleaning pump and the historical benchmark value corresponding to the theoretical flow rate;

[0025] When a preset allowable deviation threshold exists, it is determined that the cleaning pump or regulating valve has experienced performance degradation.

[0026] Based on the degree of deviation between the actual output pressure and the actual input power of the cleaning pump, the corresponding relationship between the operating parameters of the cleaning pump, the regulating valve opening parameters, and the cleaning fluid flow rate is dynamically corrected.

[0027] Based on the corrected relationship between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid, the reference value of the cleaning fluid flow rate is calculated.

[0028] Furthermore, before conducting a consistency assessment, the process also includes identifying the operating mode of the cleaning process; and adjusting the preset flow deviation threshold and the preset threshold range based on the identified operating mode of the cleaning process.

[0029] Furthermore, conducting a conformity assessment also includes:

[0030] Identify the operational characteristics of the cleaning process;

[0031] Based on the characteristics of the operating scenario of the cleaning process, obtain the dynamic fluctuation range corresponding to the characteristics of the operating scenario;

[0032] Collect cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters;

[0033] Time-series analysis was performed on the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters to extract the dynamic characteristics of the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters.

[0034] The extracted dynamic features and dynamic fluctuation range are compared.

[0035] Based on the comparison results, calculate the comprehensive anomaly index;

[0036] When the comprehensive anomaly index continues to exceed the preset anomaly threshold, the cleaning fluid flow measurement signal is deemed unreliable.

[0037] Furthermore, the steps of performing time-series analysis on the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters to extract the dynamic characteristics of these parameters include:

[0038] The system monitors the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters in real time to identify whether there are short-term or instantaneous spikes or punctures in these parameters.

[0039] When instantaneous spikes or burrs are detected, the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter and pipeline pressure parameter affected by the instantaneous spikes or burrs are filtered to eliminate or weaken the influence of the instantaneous spikes or burrs.

[0040] During the filtering process of cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters affected by instantaneous spikes or burrs, the dynamic feature extraction of the affected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters is paused or adjusted.

[0041] After filtering and removing the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch, the dynamic characteristics of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch are restored and extracted. Then, time series analysis is performed on the processed cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch to extract the dynamic characteristics of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch.

[0042] For cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters that are not affected by instantaneous spikes or glitch, continuous time-series analysis is performed to extract the dynamic characteristics of these parameters.

[0043] By fusing all dynamic features, a set of dynamic features reflecting the true operating state of the system is obtained.

[0044] Furthermore, when comparing the extracted dynamic features with the dynamic fluctuation range, the triggered conditions include:

[0045] When a transitional condition occurs, the comparison of the extracted dynamic features and dynamic fluctuation range is paused.

[0046] After the transition condition ends, once the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter have stabilized, the comparison of the extracted dynamic characteristics and dynamic fluctuation range will resume.

[0047] Furthermore, based on the comparison results, the steps for calculating the comprehensive anomaly index include:

[0048] Identify the degree of deviation of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter in the comparison results;

[0049] Based on the degree of deviation and the correlation between the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in the historical composite fault modes, the weights of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in the calculation of the comprehensive anomaly index are dynamically adjusted.

[0050] The deviations after weighting are aggregated to obtain a comprehensive anomaly index.

[0051] Furthermore, the steps for identifying the degree of deviation include:

[0052] Obtain the comparison results;

[0053] The deviation in the results is normalized to obtain the normalized deviation value.

[0054] Based on the normalized deviation value, identify the deviation direction of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter;

[0055] When there are multiple deviations in opposite directions, the weighted difference of the normalized deviation values ​​of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in opposite directions is calculated to obtain the offsetting effect.

[0056] Based on the amount of the offsetting effect, adjust the deviation of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter to obtain the adjusted deviation.

[0057] The deviation direction and the adjusted deviation magnitude are considered as the degree of deviation.

[0058] This application also discloses a flow control system for cleaning liquid in a wine fermentation valve array, applied to the above-mentioned flow control method for cleaning liquid in a wine fermentation valve array. The system includes:

[0059] The data acquisition module is used to acquire cleaning fluid flow measurement signals, cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters.

[0060] The calculation module is used to establish the correspondence between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid, and to calculate the reference value of the cleaning fluid flow rate based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve.

[0061] The evaluation module is used to evaluate the consistency of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameter. When the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds the preset flow deviation threshold, and the regulating valve opening parameter or the pipeline pressure parameter exceeds the preset threshold range, the cleaning fluid flow measurement signal is determined to be unreliable.

[0062] The control switching module switches the control mode when it determines that the cleaning fluid flow measurement signal is unreliable, and controls the cleaning fluid flow based on the cleaning fluid flow reference value in the switched control mode.

[0063] The alarm module is used to issue maintenance alarms.

[0064] This system enables intelligent control of the cleaning fluid flow rate, effectively identifies and handles unreliable flow measurement signals, avoids control deviations caused by sensor failures or changes in cleaning fluid characteristics, thereby protecting the valve array system, extending equipment life, and issuing timely maintenance alarms to reduce maintenance costs.

[0065] This application effectively solves the technical problem in the prior art where the inaccurate flow meter measurement caused by the recycling of cleaning fluid leads to the actual flow rate of the cleaning fluid being higher than the design value, causing hidden erosion and accelerated wear of the valve array system, which is difficult to identify and diagnose.

[0066] To further understand the features and technical content of the present invention, please refer to the following detailed description and accompanying drawings. However, the drawings provided are for reference and illustration only and are not intended to limit the present invention. Attached Figure Description

[0067] Figure 1 This is a flowchart of a method for controlling the flow rate of a cleaning solution for a wine fermentation valve array according to the present invention.

[0068] Figure 2 This is a schematic diagram of the structure of a wine fermentation valve array cleaning liquid flow control system according to the present invention. Detailed Implementation

[0069] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can understand the advantages and effects of the present invention from the content disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. Furthermore, the accompanying drawings of the present invention are for simple illustrative purposes only and are not depictions of actual dimensions; this is stated in advance. The following embodiments will further describe the relevant technical content of the present invention in detail, but the disclosed content is not intended to limit the scope of protection of the present invention.

[0070] This embodiment provides a method and system for controlling the flow rate of cleaning liquid in a wine fermentation valve array, combined with... Figure 1 and Figure 2 As shown.

[0071] refer to Figure 1This application proposes a method for controlling the flow rate of cleaning liquid in a wine fermentation valve array, the method comprising the following steps:

[0072] Collect cleaning fluid flow measurement signals, cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters;

[0073] Establish the correspondence between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid; and calculate the reference value of the cleaning fluid flow rate based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve.

[0074] The consistency of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameter is evaluated. When the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds the preset flow deviation threshold, and the regulating valve opening parameter or the pipeline pressure parameter exceeds the preset threshold range, the cleaning fluid flow measurement signal is determined to be unreliable.

[0075] When the cleaning fluid flow measurement signal is determined to be unreliable, the control mode is switched, and the cleaning fluid flow is controlled according to the cleaning fluid flow reference value in the switched control mode.

[0076] A maintenance alert has been issued.

[0077] The "cleaning fluid flow measurement signal" mentioned in this application refers to an electrical or digital signal output by a flow meter (e.g., an electromagnetic flow meter) that reflects the instantaneous flow rate of the cleaning fluid. "Cleaning pump operating parameters" may include the pump's speed, motor current, voltage, frequency, etc., which indirectly reflect the actual working state and output capacity of the cleaning pump. "Control valve opening parameters" refer to the degree of opening of the control valve, usually expressed as a percentage, which directly affects the flow resistance of the cleaning fluid in the pipeline. "Pipeline pressure parameters" refer to the pressure value in the cleaning pipeline, such as the cleaning pump outlet pressure or valve array inlet pressure; this parameter is closely related to flow rate and resistance. "Cleaning fluid flow reference value" is the expected flow rate value calculated through a model based on the theoretical or historical operating characteristics of the cleaning pump and control valve, serving as a reference for the flow measurement signal. "Consistency assessment" refers to a comprehensive analysis of multiple related parameters to determine whether there are logical contradictions or abnormal deviations among them, thereby assessing the reliability of the measurement signal. "Preset flow deviation threshold" refers to the standard for judging whether the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value is too large. "Preset threshold range" refers to the normal fluctuation range of the regulating valve opening parameter or pipeline pressure parameter. Exceeding this range may indicate an anomaly in the system. "Control mode" refers to the strategy or data source on which the system performs flow control. For example, it can be a closed-loop control mode based on flow meter measurements, or an open-loop or semi-open-loop control mode based on calculated reference values.

[0078] Specifically, the cleaning fluid flow control method of this application includes the following steps:

[0079] First, the cleaning fluid flow rate signal can be acquired using flow meters installed on the cleaning pipeline, such as electromagnetic flow meters, ultrasonic flow meters, or vortex flow meters. These flow meters convert the instantaneous flow rate of the cleaning fluid into electrical or digital signals and transmit them to the control system. The cleaning pump operating parameters can be acquired by installing current sensors, voltage sensors, or speed sensors on the cleaning pump motor. These sensors monitor the operating status of the cleaning pump in real time. The regulating valve opening parameters can be acquired using the regulating valve's built-in opening sensor or position feedback device. This sensor converts the actual opening degree of the valve into an electrical signal. Pipeline pressure parameters can be acquired by installing pressure sensors at key locations in the cleaning pipeline (such as the cleaning pump outlet, valve array inlet, or critical pipe sections). These sensors monitor pressure changes within the pipeline in real time. All acquired signals are transmitted to the central control unit or data processing module for further processing.

[0080] Secondly, the correspondence between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid is established, and a reference value for the cleaning fluid flow rate is calculated based on these parameters. This correspondence can be established in several ways. For example, regression analysis can be performed using historical operating data to establish a multinomial or neural network model, using parameters such as the pump speed, motor current, and regulating valve opening as inputs, and the cleaning fluid flow rate as the output. Another approach is to establish a physical model based on fluid mechanics principles. For instance, the theoretical relationship between flow rate and these parameters can be derived from the pump performance curve, pipeline resistance characteristics, and regulating valve flow characteristic curve. In practical applications, the system can be calibrated periodically by measuring the actual flow rate under different operating conditions and recording the corresponding cleaning pump operating parameters and regulating valve opening parameters, thereby calibrating or optimizing the established correspondence. Once the correspondence is established, the system can calculate the current reference value for the cleaning fluid flow rate based on the real-time collected cleaning pump operating parameters and regulating valve opening parameters using this model.

[0081] Next, a consistency assessment is performed on the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameters. When the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds a preset flow deviation threshold, and the regulating valve opening parameter or pipeline pressure parameter exceeds a preset threshold range, the cleaning fluid flow measurement signal is deemed unreliable. Consistency assessment can be implemented using various algorithms. For example, a rule-based expert system can be used, with a set of preset logical judgment rules. When a specific combination of conditions is met, an unreliable determination is triggered. Specifically, the system continuously monitors the real-time deviation between the cleaning fluid flow measurement signal and the calculated cleaning fluid flow reference value. If this deviation continuously or momentarily exceeds the preset flow deviation threshold (e.g., exceeding the set flow rate by 5%), it indicates that the measurement signal may be abnormal. Simultaneously, the system also monitors the regulating valve opening parameter and pipeline pressure parameter. For example, when the flow measurement signal is low, but the regulating valve opening has reached its maximum value (e.g., above 95%), or the pipeline pressure abnormally increases or decreases (e.g., exceeding the normal operating pressure range by 10%), these situations may indicate inaccurate flow meter measurement. When multiple conditions are met simultaneously, the system will determine that the cleaning fluid flow measurement signal is unreliable.

[0082] Finally, when the cleaning fluid flow measurement signal is determined to be unreliable, the control mode is switched. In the switched control mode, the cleaning fluid flow is controlled based on the cleaning fluid flow reference value, and a maintenance alarm is issued. Once the system determines that the cleaning fluid flow measurement signal is unreliable, the control system will immediately switch its control mode to avoid problems caused by control based on erroneous measurements (such as over-flushing). In the switched control mode, the system no longer relies on the unreliable cleaning fluid flow measurement signal, but instead performs flow control based on the previously calculated cleaning fluid flow reference value. For example, it can switch from closed-loop control to open-loop control or model-based predictive control, adjusting the cleaning pump speed or the opening of the regulating valve to make the actual flow as close as possible to the cleaning fluid flow reference value. At the same time, the system will immediately issue a maintenance alarm, such as through audible and visual alarms, sending messages to the operator interface, or sending notifications to maintenance personnel via the network, indicating that the flow measurement system may have a fault or abnormality and requires inspection and maintenance.

[0083] The core innovation of this application lies in its integration of not only the cleaning fluid flow rate measurement signal but also the cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters. It establishes a correlation between these parameters and the cleaning fluid flow rate, enabling the calculation of a reference value for the cleaning fluid flow rate. This reference value is independent of the flow meter's measurement, providing a crucial benchmark for assessing the flow meter's reliability. More importantly, this application conducts a consistency evaluation of the cleaning fluid flow rate measurement signal, the reference value, and the pipeline pressure parameters. When the measured signal deviates significantly from the reference value, and the regulating valve opening or pipeline pressure also shows abnormalities, it can intelligently determine that the flow rate measurement signal is unreliable. This multi-dimensional, multi-parameter cross-validation mechanism significantly improves the system's ability to identify flow measurement anomalies, avoiding misjudgments caused by a single sensor failure.

[0084] When the measurement signal is deemed unreliable, this application can promptly switch the control mode to control based on a more reliable cleaning fluid flow rate reference value, thereby avoiding over-flushing caused by control based on erroneous measurement values. Simultaneously, the system will issue a maintenance alarm, reminding maintenance personnel to intervene promptly, fundamentally solving the problem of hidden and difficult-to-diagnose issues in traditional methods. Compared to existing technologies, this application's method effectively prevents "stable over-flow rate" scouring caused by flowmeter contamination, extends the service life of the valve array system, reduces maintenance costs, and improves the reliability and safety of the cleaning process. This proactive ability to identify and respond to measurement anomalies is not found in existing technologies, demonstrating the significant advancements in intelligent control and fault diagnosis in this application.

[0085] This application further proposes steps for conducting a conformity assessment, including:

[0086] Obtain the opening command signal of the control valve;

[0087] The opening command signal of the control valve is subjected to high-pass filtering to obtain the rapidly changing components of the command signal;

[0088] Calculate the root mean square value of the rapidly changing component of the command signal to obtain the quantitative index of the rapidly changing component of the command signal.

[0089] Based on the quantification index of the rapidly changing components of the command signal, the proportional gain and derivative gain of the controller used for consistency evaluation are dynamically adjusted.

[0090] When the quantization index of the rapidly changing components of the command signal exceeds the preset upper limit threshold, the proportional gain and derivative gain of the controller are reduced.

[0091] When the quantization index of the rapidly changing component of the command signal remains below the preset upper limit threshold, the proportional gain and derivative gain of the controller are restored.

[0092] Based on the adjusted controller, a consistency assessment is performed on the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameters.

[0093] Specifically, before conducting the consistency assessment, the valve opening command signal is first acquired. This command signal reflects the control system's intention to adjust the valve opening, and its rate and magnitude of change indirectly reflect the current system's dynamic characteristics. To identify the rapidly changing components in the command signal, a high-pass filter is applied. The high-pass filter effectively removes low-frequency components while retaining high-frequency components, thus revealing the rapidly changing component of the command signal. This rapidly changing component represents the instantaneous rate of change or fluctuation intensity of the valve opening command.

[0094] Furthermore, to quantify the degree of rapid change in the command signal, the root mean square (RMS) value of the rapidly changing component of the command signal is calculated. The RMS value reflects the effective value or energy of the signal, and therefore can be used as a quantification index of the rapidly changing component of the command signal. The higher this quantification index, the more drastic the change in the valve opening command, and the more dynamic the system is under operating conditions.

[0095] Based on this quantitative index, the proportional gain and derivative gain of the controller used for consistency evaluation are dynamically adjusted. Specifically, when the quantitative index of the rapidly changing component of the command signal exceeds a preset upper threshold, it indicates that the system is undergoing significant dynamic changes. At this time, the proportional gain and derivative gain of the controller are reduced. The purpose of reducing the gain is to decrease the controller's sensitivity to transient fluctuations and avoid misjudging the unreliability of the cleaning fluid flow measurement signal due to instantaneous parameter deviations during the normal dynamic response of the system. Conversely, when the quantitative index of the rapidly changing component of the command signal remains below the preset upper threshold, it indicates that the system is stabilizing or in steady-state operation. At this time, the proportional gain and derivative gain of the controller are restored. The purpose of restoring the gain is to ensure that the controller can maintain sufficient response speed and evaluation accuracy under stable operating conditions.

[0096] In some preferred embodiments, it is assumed that precise flow control of a certain cleaning loop is required during the cleaning process of the wine fermentation valve array. When the cleaning pump starts or the regulating valve is adjusted to a large degree, the opening command signal of the regulating valve will show a significant rapid change. The system will acquire this command signal in real time and extract its rapidly changing component through high-pass filtering. For example, if the command signal rapidly increases from 20% opening to 80% opening in a short period of time, a large rapidly changing component will be obtained after high-pass filtering. At this time, the root mean square value of this rapidly changing component is calculated. If the value exceeds a preset upper limit threshold (for example, indicating that the valve opening change rate exceeds 10% / second), the system will immediately reduce the proportional gain and derivative gain of the PID controller used for consistency evaluation. For example, the proportional gain is reduced from 0.5 to 0.2, and the derivative gain is reduced from 0.1 to 0.05. In the state of reduced gain, even if there is a brief deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value due to the transient response of the system, the controller will not immediately determine that the measurement signal is unreliable. Once the valve opening stabilizes, the root mean square value of the rapidly changing component of the command signal will remain below the upper limit threshold. At this point, the controller gain will return to its initial value to ensure high-precision reliability assessment of the flow measurement signal under stable operating conditions.

[0097] This application further proposes a step for estimating the reference value of the cleaning fluid flow rate, which includes:

[0098] Periodically perform a performance self-calibration procedure on the cleaning pump and control valve;

[0099] At the preset operating point, measure the actual output pressure and actual input power of the cleaning pump;

[0100] The theoretical flow rate value is obtained based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve;

[0101] Determine whether there is a preset allowable deviation threshold between the actual output pressure or actual input power of the cleaning pump and the historical benchmark value corresponding to the theoretical flow rate;

[0102] When a preset allowable deviation threshold exists, it is determined that the cleaning pump or regulating valve has experienced performance degradation.

[0103] Based on the degree of deviation between the actual output pressure and the actual input power of the cleaning pump, the corresponding relationship between the operating parameters of the cleaning pump, the regulating valve opening parameters, and the cleaning fluid flow rate is dynamically corrected.

[0104] Based on the corrected relationship between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid, the reference value of the cleaning fluid flow rate is calculated.

[0105] Specifically, the periodic performance self-calibration procedure for the cleaning pump and control valve refers to the system automatically executing a series of tests at preset time intervals or under specific operating conditions, such as before each cleaning task or at regular intervals, to evaluate the current performance status of the cleaning pump and control valve. At preset operating points, such as a stable state where the cleaning pump is running at a specific speed and the control valve is open at a specific degree, the actual output pressure and actual input power of the cleaning pump are measured. These parameters are key indicators reflecting the cleaning pump's efficiency and energy consumption. The theoretical flow rate can be understood as the expected flow rate based on the operating and opening parameters of the cleaning pump and control valve under ideal or initial performance conditions. The historical baseline value refers to the reference values ​​of output pressure and input power measured at the same operating point under normal, non-attenuating conditions. The system determines whether the actual output pressure or actual input power of the cleaning pump differs from the historical baseline value corresponding to the theoretical flow rate value within a preset allowable deviation threshold. When the deviation between the actual measured value and the historical baseline value exceeds the preset allowable deviation threshold, it is determined that the cleaning pump or control valve has experienced performance degradation. Once performance degradation is detected, the system dynamically adjusts the relationship between the cleaning pump's operating parameters, regulating valve opening parameters, and cleaning fluid flow rate based on the deviation between the actual output pressure and actual input power of the cleaning pump. This adjustment can be based on a preset adjustment model, lookup table, or adaptive algorithm to ensure that the adjusted relationship more accurately reflects the actual flow characteristics of the equipment under its current performance state. Finally, based on the adjusted relationship between the cleaning pump's operating parameters, regulating valve opening parameters, and cleaning fluid flow rate, the system calculates a more accurate reference value for the cleaning fluid flow rate.

[0106] In some preferred embodiments, it is assumed that during the cleaning process of the wine fermentation valve array, the cleaning pump experiences slight wear on its internal impeller after prolonged operation, resulting in a slightly lower output pressure than the initial design value at the same rotational speed and opening degree. Without the solution described in this application, the system would still calculate the reference value of the cleaning fluid flow rate based on the initially established correspondence, which might lead to a calculated value higher than the actual flow rate, thus affecting the accuracy of flow control.

[0107] This application further proposes an optimization scheme aimed at improving the accuracy and adaptability of conformity assessment.

[0108] Before conducting a consistency assessment, the process also includes identifying the operating mode of the cleaning process; based on the identified operating mode of the cleaning process, the preset flow deviation threshold and the preset threshold range are adjusted.

[0109] Specifically, identifying the operating mode of the cleaning process refers to the system's ability to automatically or manually identify the specific working condition or stage of the current cleaning task. For example, operating modes may include pre-wash mode, main wash mode, disinfection mode, and rinsing mode. Each operating mode may correspond to different cleaning fluid flow rates, pressures, and valve opening characteristics. Identifying the operating mode can be achieved in various ways, such as based on preset cleaning procedure steps, by monitoring the type, temperature, or pH value of the cleaning fluid, or by analyzing cleaning pump operating parameters and regulating valve opening parameters.

[0110] This process involves adjusting preset flow deviation thresholds and ranges based on the identified operating mode of the cleaning process. This can be understood as dynamically selecting or calculating the appropriate flow deviation threshold and range for the currently identified operating mode. For example, a specific set of flow deviation thresholds and ranges can be pre-configured for each operating mode and applied when the corresponding mode is identified. Alternatively, an adaptive algorithm can be used to calculate the optimal threshold in real time based on historical data and the characteristics of the current operating mode. The preset flow deviation threshold is used to determine whether the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value is acceptable, while the preset threshold range is used to determine whether the regulating valve opening parameter or pipeline pressure parameter is within the normal operating range. This dynamic adjustment ensures that the sensitivity and accuracy of the consistency assessment match the current actual operating conditions.

[0111] This application's solution effectively solves the problem of misjudgment and missed judgment that may occur with fixed thresholds in the basic solution by first identifying the operating mode of the cleaning process before conducting a consistency assessment, and then dynamically adjusting the preset flow deviation threshold and preset threshold range based on the mode. Because the normal fluctuation range of system parameters differs under different operating modes, adaptive adjustment of the threshold allows the consistency assessment to more accurately reflect the actual situation under the current operating conditions. For example, in some cleaning stages, the system may allow for large flow fluctuations; in this case, using a strict fixed threshold may frequently trigger false alarms. Conversely, in stages requiring precise control, using a lenient fixed threshold may fail to detect subtle anomalies in a timely manner. By identifying the operating mode and adjusting the threshold, the system can adopt the most appropriate evaluation standard under different operating conditions according to actual needs, thereby significantly improving the accuracy and robustness of the reliability determination of the cleaning fluid flow measurement signal.

[0112] In some preferred embodiments, it is assumed that the cleaning process of the wine fermentation valve array is divided into three main operating modes: pre-rinse mode, main cleaning mode, and disinfection mode.

[0113] In pre-flushing mode, the main purpose is to quickly flush out residues in the pipeline. The accuracy requirement for flow rate is relatively low, and pipeline pressure fluctuations may be large. At this time, the system recognizes that it is in pre-flushing mode and can automatically set the preset flow rate deviation threshold to a relatively lenient value (e.g., ±10%), and set the preset threshold ranges for regulating valve opening parameters and pipeline pressure parameters to a wider range to allow for larger normal fluctuations.

[0114] When entering the main cleaning mode, precise control of the cleaning fluid flow rate is required to ensure cleaning effectiveness, while the pipeline pressure should remain relatively stable. After the system recognizes the main cleaning mode, it will immediately adjust the preset flow deviation threshold to a more stringent value (e.g., ±3%), and tighten the preset threshold ranges of the regulating valve opening parameters and pipeline pressure parameters to improve sensitivity to minor anomalies.

[0115] In disinfection mode, specific flow rates and pressures may be required to ensure the effective action of the disinfectant, and the system typically needs to operate in a very stable state. In this mode, the system will adjust the preset flow deviation threshold and preset threshold range to the most stringent level to ensure that any minute deviations are detected and addressed promptly.

[0116] In this way, the system can intelligently adjust the consistency evaluation criteria according to the current actual operating mode, thereby significantly improving the accuracy and reliability of the confidence judgment of the cleaning fluid flow measurement signal and avoiding misjudgment or omission due to threshold mismatch under different operating conditions.

[0117] This application further proposes a more refined method for evaluating the reliability of cleaning fluid flow measurement signals. By introducing the identification of the characteristics of the cleaning process operation scenario and dynamic feature analysis, the reliability of the signal can be judged more accurately.

[0118] The conformity assessment also includes:

[0119] Identify the operational characteristics of the cleaning process;

[0120] Based on the characteristics of the operating scenario of the cleaning process, obtain the dynamic fluctuation range corresponding to the characteristics of the operating scenario;

[0121] Collect cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters;

[0122] Time-series analysis was performed on the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters to extract the dynamic characteristics of the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters.

[0123] The extracted dynamic features and dynamic fluctuation range are compared.

[0124] Based on the comparison results, calculate the comprehensive anomaly index;

[0125] When the comprehensive anomaly index continues to exceed the preset anomaly threshold, the cleaning fluid flow measurement signal is deemed unreliable.

[0126] Specifically, identifying the operational scenario characteristics of the cleaning process refers to the system's ability to determine the specific operational situation or stage of the current cleaning process based on the changing trends of current control commands, equipment status (e.g., the opening or closing of specific valves, the start-up or shutdown of cleaning pumps, or changes in pump speed), and process parameters (e.g., rapid increases or decreases in pipeline pressure). These operational scenario characteristics are typically more granular than the operational mode. For example, the operational mode of "cleaning cycle" can be further subdivided into specific operational scenarios such as "valve array switching," "pump start-up," "pipeline filling," and "pipeline evacuation."

[0127] Specifically, obtaining the dynamic fluctuation range corresponding to the operational scenario characteristics based on the cleaning process means that for each identified operational scenario characteristic, the system will acquire the expected dynamic change range or pattern of key parameters such as the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter under that scenario. These dynamic fluctuation ranges can be predetermined and stored through historical operational data analysis, system modeling, or expert experience. For example, in the "valve array switching" scenario, flow and pressure may experience brief fluctuations, and the amplitude, duration, and rate of change of these fluctuations all have specific expected ranges.

[0128] In practical applications, time-series analysis of the collected cleaning fluid flow measurement signals, reference values, valve opening parameters, and pipeline pressure parameters is performed to extract their dynamic characteristics. This involves in-depth analysis of these time-varying signals to reveal their inherent patterns and features. Dynamic characteristics may include, but are not limited to, the signal's instantaneous value, rate of change (first derivative), acceleration (second derivative), mean, variance, peak value, trough value, frequency components, trend, and periodicity. By performing time-series analysis on these parameters, a more comprehensive understanding of the signal's dynamic behavior in specific scenarios can be achieved.

[0129] Furthermore, the extracted dynamic features and dynamic fluctuation ranges are compared to determine whether the currently extracted dynamic features fall within the expected dynamic fluctuation range of the corresponding operating scenario. This comparison can employ various methods, such as comparing the current signal's rate of change with historical normal fluctuation rates, or matching the signal's spectral characteristics with a preset normal spectral pattern. Based on the comparison results, a comprehensive anomaly index is calculated. This index is a quantitative indicator used to comprehensively reflect the overall degree of anomaly of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter under the current operating scenario. The calculation of this index can consider the degree of deviation, direction of deviation, and correlation between each parameter. When the comprehensive anomaly index continuously exceeds a preset anomaly threshold, the cleaning fluid flow measurement signal is deemed unreliable. "Continuously exceeding" means that the abnormal state is not instantaneous but has persisted for a period of time. This helps avoid misjudgments caused by brief interference or normal transient fluctuations, thereby improving the robustness of the judgment.

[0130] Specifically, in complex cleaning processes, system behavior is not always stable, especially during transitional conditions such as valve switching and pump start-up / stop, where parameters like flow rate and pressure exhibit specific dynamic changes. If relying solely on static thresholds within the operating mode, these normal dynamic fluctuations may be misjudged as abnormal, or genuine sensor malfunctions may be masked by these fluctuations. By identifying more granular operating scenario characteristics, the system can establish a more precise dynamic fluctuation range for each specific scenario, reflecting the expected dynamic behavior of parameters under that scenario. Subsequently, time-series analysis of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameters, and pipeline pressure parameters allows for the extraction of dynamic features such as instantaneous rate of change, trend, and frequency. These features reflect the true state of the signal more accurately than a single instantaneous value. Comparing these extracted dynamic features with the dynamic fluctuation range of the corresponding scenario allows for a more accurate determination of whether the signal has deviated from its normal dynamic trajectory under that scenario. Therefore, by calculating the comprehensive anomaly index and combining it with the "continuous exceedance" criterion, this solution can effectively distinguish between transient fluctuations caused by changes in normal operating conditions and real anomalies caused by sensor failures, significantly improving the accuracy and robustness of the reliability assessment of cleaning fluid flow measurement signals.

[0131] In some preferred embodiments, it is assumed that during the valve array cleaning process in wine fermentation, the system switches from a "filling" mode to a "circulating cleaning" mode, which includes a "rapid valve array switching" operational scenario. In conventional evaluation methods, a flow deviation threshold may only be set based on the "circulating cleaning" mode. When the valve array switches rapidly, the cleaning fluid flow measurement signal and pipeline pressure parameters may experience a brief but significant instantaneous fluctuation. If this fluctuation exceeds the preset static flow deviation threshold, even if this is an expected phenomenon during normal switching, it may be misjudged as an unreliable cleaning fluid flow measurement signal. However, in the solution of this application, the system first identifies the "rapid valve array switching" operational scenario. Based on historical data or a preset model, the system knows that under this scenario, the cleaning fluid flow measurement signal and pipeline pressure parameters will have a specific dynamic fluctuation range. For example, the flow rate may decrease by 20% within 0.5 seconds and then quickly recover, while the pressure may increase by 15% within 0.3 seconds and then fall back. The system then performs time-series analysis on the collected cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameters, and pipeline pressure parameters to extract their dynamic characteristics, such as calculating their instantaneous rate of change and fluctuation duration. By comparing the extracted dynamic characteristics (e.g., an actual flow rate decrease of 18% lasting 0.4 seconds) with the preset dynamic fluctuation range, the system finds that these dynamic characteristics fall within the expected range. Even if the instantaneous deviation may be large, because its dynamic behavior conforms to the characteristics of this scenario, the calculated comprehensive anomaly index will remain below the preset anomaly threshold. Therefore, the system can accurately determine that the cleaning fluid flow measurement signal is reliable at this moment, avoiding unnecessary control mode switching and maintenance alarms, thereby improving the system's intelligence level and operational robustness. Conversely, if the actual flow rate decrease reaches 50% and lasts for 2 seconds, significantly exceeding the dynamic fluctuation range of this scenario, the comprehensive anomaly index will continuously exceed the anomaly threshold, and the system will accurately determine that the cleaning fluid flow measurement signal is unreliable and take appropriate measures in a timely manner.

[0132] The steps described above for performing time-series analysis on the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters, and extracting the dynamic characteristics of these parameters, include:

[0133] The system monitors the collected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters in real time to identify whether there are short-term or instantaneous spikes or punctures in these parameters.

[0134] When instantaneous spikes or burrs are detected, the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter and pipeline pressure parameter affected by the instantaneous spikes or burrs are filtered to eliminate or weaken the influence of the instantaneous spikes or burrs.

[0135] During the filtering process of cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters affected by instantaneous spikes or burrs, the dynamic feature extraction of the affected cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters is paused or adjusted.

[0136] After filtering and removing the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch, the dynamic characteristics of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch are restored and extracted. Then, time series analysis is performed on the processed cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch to extract the dynamic characteristics of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter affected by instantaneous spikes or glitch.

[0137] For cleaning fluid flow measurement signals, cleaning fluid flow reference values, regulating valve opening parameters, and pipeline pressure parameters that are not affected by instantaneous spikes or glitch, continuous time-series analysis is performed to extract the dynamic characteristics of these parameters.

[0138] By fusing all dynamic features, a set of dynamic features reflecting the true operating state of the system is obtained.

[0139] "Short-term, instantaneous spikes or glitches" refer to abnormal signal points or small-scale, violent fluctuations that occur outside of normal signal fluctuations, caused by sensor noise, electromagnetic interference, instantaneous pressure shocks, or data transmission errors. These fluctuations are extremely short in duration but may have large amplitudes. Real-time monitoring can be achieved by setting dynamic thresholds, using statistical methods (such as Z-score, IQR), or employing machine learning anomaly detection algorithms.

[0140] Specifically, filtering can employ various digital signal processing techniques, such as median filtering, mean filtering, Kalman filtering, and wavelet denoising. The aim is to smooth the signal and remove or significantly reduce the impact of these transient outliers on the overall signal shape and subsequent feature extraction.

[0141] In practical applications, "pause or adjust" means that at critical stages of signal processing, to avoid misinterpreting transient responses or imperfectly smoothed signals during filtering as dynamic features, the extraction of dynamic features from that part of the signal can be temporarily stopped, or a more conservative and robust feature extraction algorithm can be adopted. For example, the calculation of features sensitive to instantaneous changes, such as the derivative and variance of the signal, can be paused.

[0142] Once the filtering process is complete and the signal returns to a stable or acceptable state, the normal dynamic feature extraction process can resume. At this point, time-series analysis of the filtered signal can extract more realistic dynamic features with less noise interference.

[0143] For signal segments where no transient spikes or glitches are detected, no additional filtering is required, and routine time-series analysis and dynamic feature extraction can continue to ensure data continuity and integrity.

[0144] "Fusion" refers to integrating processed (filtered and extracted) and unprocessed (directly extracted) dynamic features. This can be achieved through weighted averaging, feature vector concatenation, or machine learning-based feature fusion methods to form a comprehensive and accurate feature set that reflects the system's true dynamic behavior. The goal is to ensure that critical dynamic information about the system under normal operating conditions is not lost while removing noise.

[0145] Specifically, when the system detects short-term or instantaneous spikes or glitches in the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameters, and pipeline pressure parameters, it immediately filters the affected signals to eliminate or significantly reduce the impact of these instantaneous anomalies. During this filtering process, to avoid misinterpreting transient responses or imperfectly smoothed signals as dynamic features, the system pauses or adjusts the extraction of dynamic features from the affected signals. Once the filtering is complete and the signal returns to a stable state, normal dynamic feature extraction resumes. For signals unaffected by instantaneous spikes or glitches, time-series analysis continues. Finally, all extracted dynamic features are fused, including features generated by both processed and unprocessed signals, ensuring that the obtained dynamic feature set more accurately and robustly reflects the system's true operating state and avoids misinterpretations caused by instantaneous interference.

[0146] In some preferred embodiments, it is assumed that during the cleaning process of the wine fermentation valve array, due to the instantaneous fluctuations of the on-site electromagnetic environment, the cleaning liquid flow measurement signal will have a spike with an extremely short duration but an abnormally high amplitude at a certain moment.

[0147] First, the system will monitor this abnormal instantaneous spike in real time. Once the spike is identified, the system will immediately perform median filtering on the affected cleaning fluid flow measurement signal to eliminate or weaken the impact of the spike and restore it to the normal signal range.

[0148] During the brief period of median filtering, the system will pause the extraction of dynamic characteristics (e.g., flow rate of change, flow fluctuation amplitude, etc.) of the flow measurement signal to avoid misjudging transient responses or residual noise that may occur during the filtering process as true dynamic characteristics.

[0149] After the filtering process is completed, the cleaning fluid flow measurement signal becomes smooth and stable. The system will then resume the dynamic feature extraction of the signal and perform time-series analysis on the processed signal to extract its true dynamic features.

[0150] Meanwhile, for signals such as cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters that are not affected by instantaneous spikes or glitches, the system will continuously perform routine time-series analysis and dynamic feature extraction.

[0151] Finally, all extracted dynamic features, including those from the filtered flow measurement signal and other unaffected signals, are fused. For example, a comprehensive set of dynamic features can be formed by concatenating feature vectors. This set will more accurately reflect the actual operating state of the cleaning process, thus providing a high-quality data foundation for subsequent consistency assessment of the cleaning fluid flow measurement signal, effectively avoiding misjudgments caused by transient interference, and ensuring the stability and reliability of the cleaning process.

[0152] This application further proposes that when comparing the extracted dynamic features with the dynamic fluctuation range, the triggered conditions include:

[0153] When a transitional condition occurs, the comparison of the extracted dynamic features and dynamic fluctuation range is paused.

[0154] After the transition condition ends, once the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter have stabilized, the comparison of the extracted dynamic characteristics and dynamic fluctuation range will resume.

[0155] Specifically, transitional operating conditions refer to the process by which a system changes from one stable operating state to another, such as starting or stopping the cleaning pump, switching cleaning modes, or significantly adjusting valve openings. Under these conditions, the system's physical parameters, such as the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameters, and pipeline pressure parameters, will undergo significant, non-linear changes. Pausing the comparison of extracted dynamic features and dynamic fluctuation ranges means temporarily halting the execution of reliability assessment logic based on dynamic feature comparison during this unstable state. The purpose is to avoid misjudging normal system dynamic responses under transitional operating conditions as sensor malfunctions or system anomalies.

[0156] After the transition period ends, once the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter stabilize, the comparison of extracted dynamic characteristics and dynamic fluctuation range resumes. Here, "stable" can be understood as the fluctuation amplitude of the above parameters converging within a preset stability threshold range, and the rate of change being lower than a preset stability rate threshold. For example, the stability of the system can be determined by monitoring the changes in the mean, variance, or rate of change of parameters within a certain time window. Once the system stabilizes, its operating state is considered to have returned to normal, and resuming the comparison operation at this point ensures the accuracy and effectiveness of the reliability assessment.

[0157] This application further proposes the following steps for calculating the comprehensive anomaly index based on the comparison results:

[0158] Identify the degree of deviation of the cleaning fluid flow measurement signal, cleaning fluid flow reference value, regulating valve opening parameter, and pipeline pressure parameter in the comparison results;

[0159] Based on the degree of deviation and the correlation between the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in the historical composite fault modes, the weights of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in the calculation of the comprehensive anomaly index are dynamically adjusted.

[0160] The deviations after weighting are aggregated to obtain a comprehensive anomaly index.

[0161] Specifically, after comparing the extracted dynamic features and dynamic fluctuation ranges, it is necessary to identify the degree of deviation of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter. The degree of deviation refers to the difference between the dynamic characteristics of each parameter and its corresponding dynamic fluctuation range, which can be quantified as the magnitude or direction of the deviation. For example, when the dynamic characteristics of a parameter exceed its normal fluctuation range, its degree of deviation will be identified and quantified.

[0162] Furthermore, after identifying the degree of deviation of each parameter, the weights of these parameters in the calculation of the comprehensive anomaly index are dynamically adjusted based on these deviations and the correlation between the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in historical composite failure modes. Historical composite failure modes refer to typical failure scenarios in which multiple parameters simultaneously become abnormal during the system's past operation. By analyzing this historical data, a correlation model between different parameters under specific failure modes can be established. For example, under a certain sensor drift failure mode, the deviation of the cleaning fluid flow measurement signal may be strongly correlated with the deviation of the pipeline pressure parameter. Based on this correlation, when a similar deviation pattern is detected, the weights of the relevant parameters are dynamically adjusted to more accurately reflect the impact of the current anomaly on the overall reliability of the system.

[0163] Therefore, after adjusting the weights, the deviations of each parameter are aggregated to obtain a comprehensive anomaly index. This aggregation process can employ weighted summation, weighted average, or other multivariate fusion algorithms to combine the deviations of each parameter and their adjusted weights, forming a single index that comprehensively reflects the abnormal state of the system.

[0164] In some preferred embodiments, it is assumed that during the cleaning process of the wine fermentation valve array, the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter are collected and subjected to time-series analysis to extract their respective dynamic characteristics. After comparing these dynamic characteristics with a preset dynamic fluctuation range, it is identified that the deviation of the cleaning fluid flow measurement signal is moderate, while the deviation of the pipeline pressure parameter is high. Simultaneously, by analyzing historical composite fault mode data, it is found that when the cleaning fluid flow measurement signal has a moderate deviation and the pipeline pressure parameter has a high deviation, it usually indicates a drift fault in the flow sensor. Based on this correlation, when calculating the comprehensive anomaly index, the weight of the pipeline pressure parameter is dynamically increased, while the weight of the cleaning fluid flow measurement signal may remain unchanged or be slightly decreased. Therefore, even if the deviation of the cleaning fluid flow measurement signal itself is not the highest, because of its strong correlation with the high-deviation pipeline pressure parameter in historical fault modes, the final aggregated comprehensive anomaly index will more accurately reflect the potential unreliable state of the flow sensor, thereby triggering a maintenance alarm in a timely manner.

[0165] Specifically, the steps for identifying the degree of deviation mentioned above include:

[0166] Obtain the comparison results;

[0167] The deviation in the results is normalized to obtain the normalized deviation value.

[0168] Based on the normalized deviation value, identify the deviation direction of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter;

[0169] When there are multiple deviations in opposite directions, the weighted difference of the normalized deviation values ​​of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter in opposite directions is calculated to obtain the offsetting effect.

[0170] Based on the amount of the offsetting effect, adjust the deviation of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter to obtain the adjusted deviation.

[0171] The deviation direction and the adjusted deviation magnitude are considered as the degree of deviation.

[0172] The process of acquiring comparison results involves receiving and processing the output data from the step of comparing the extracted dynamic features and dynamic fluctuation ranges. This comparison result typically includes information on the differences between the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameters, and the pipeline pressure parameters, and their corresponding dynamic fluctuation ranges.

[0173] Furthermore, the deviations in the results are normalized to obtain normalized deviation values. The purpose of normalization is to eliminate differences in the units and numerical ranges of different parameters, so that all deviations can be compared on a uniform scale. For example, methods such as min-max normalization and Z-score standardization can be used to map the original deviation values ​​to a range of 0 to 1 or -1 to 1.

[0174] Based on this, the deviation direction of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter is identified according to the normalized deviation value. The deviation direction indicates whether the current parameter value is higher or lower than its normal fluctuation range, or whether it deviates positively or negatively relative to the reference value. For example, when the cleaning fluid flow measurement signal is higher than its upper limit of the normal range, it can be identified as a positive deviation; when it is lower than the lower limit, it can be identified as a negative deviation.

[0175] When multiple deviations occur in opposite directions, the weighted difference between the normalized deviations of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter, all with opposite deviations, is calculated to obtain the cancellation effect. For example, if the cleaning fluid flow measurement signal shows a positive deviation while the pipeline pressure parameter shows a negative deviation, and the two may physically cancel each other out or be correlated, this cancellation effect can be quantified by calculating their weighted difference. The weighting coefficients can be set based on historical data analysis or expert experience to reflect the relative importance or mutual influence of different parameters in system anomalies.

[0176] Based on the offsetting effect, the deviations of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, the regulating valve opening parameter, and the pipeline pressure parameter are adjusted to obtain the adjusted deviation. The introduction of the offsetting effect aims to more accurately reflect the true abnormal state of the system and avoid underestimating or overestimating the overall abnormality due to the mutual cancellation of some parameters. For example, if the offsetting effect is large, it indicates that some deviations may be mitigated by internal system mechanisms; in this case, the deviation of the relevant parameters can be appropriately reduced.

[0177] Ultimately, the deviation direction and the adjusted deviation magnitude are used as the degree of deviation. This comprehensive consideration of the deviation direction and the adjusted deviation magnitude allows for a more complete and accurate description of the abnormal state of each parameter, providing a more reliable input for subsequent calculation of the comprehensive anomaly index.

[0178] This application's solution normalizes the original comparison results, eliminating differences in the dimensions and numerical ranges of different parameters and ensuring the fairness and accuracy of subsequent calculations. By identifying the direction of deviation for each parameter, the nature of the anomaly can be understood in more detail, such as whether the flow rate is too high or too low, or the pressure is too high or too low. Crucially, when multiple parameters in the system deviate in opposite directions, this application introduces the calculation of the offsetting effect. This mechanism can identify and quantify the potential mutual offsetting effects between different parameters. For example, a high flow sensor reading may offset a small control valve opening to some extent, ensuring that the overall system anomaly is not simply additive. By considering this offsetting effect, the deviation magnitude of each parameter can be adjusted more reasonably, thus avoiding misjudging the overall health of the system due to fluctuations in local parameters or ignoring potential complex anomalies. Therefore, the obtained deviation degree includes not only the direction of deviation but also the finely adjusted deviation magnitude, making the description of the system's abnormal state more accurate and robust.

[0179] refer to Figure 2 This application further proposes a flow control system for cleaning liquid in a wine fermentation valve array, applied to a method for controlling the flow of cleaning liquid in a wine fermentation valve array. The system includes:

[0180] The data acquisition module is used to acquire cleaning fluid flow measurement signals, cleaning pump operating parameters, regulating valve opening parameters, and pipeline pressure parameters.

[0181] The calculation module is used to establish the correspondence between the operating parameters of the cleaning pump, the opening parameters of the regulating valve, and the flow rate of the cleaning fluid, and to calculate the reference value of the cleaning fluid flow rate based on the operating parameters of the cleaning pump and the opening parameters of the regulating valve.

[0182] The evaluation module is used to evaluate the consistency of the cleaning fluid flow measurement signal, the cleaning fluid flow reference value, and the pipeline pressure parameter. When the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds the preset flow deviation threshold, and the regulating valve opening parameter or the pipeline pressure parameter exceeds the preset threshold range, the cleaning fluid flow measurement signal is determined to be unreliable.

[0183] The control switching module switches the control mode when it determines that the cleaning fluid flow measurement signal is unreliable, and controls the cleaning fluid flow based on the cleaning fluid flow reference value in the switched control mode.

[0184] The alarm module is used to issue maintenance alarms.

[0185] Specifically, the data acquisition module is configured to acquire key system operation data in real time. For example, the cleaning fluid flow rate measurement signal can be obtained by a flow meter, the cleaning pump operating parameters can include the pump speed, current or power, the regulating valve opening parameters can be provided by a valve position sensor, and the pipeline pressure parameters can be measured by a pressure sensor. The acquisition of these parameters is the foundation for subsequent data processing and control.

[0186] The calculation module is designed to build a model or lookup table to characterize the intrinsic relationship between cleaning pump operating parameters, control valve opening parameters, and cleaning fluid flow rate. Based on this correspondence, the module can calculate a theoretical reference value for the cleaning fluid flow rate according to the current cleaning pump operating parameters and control valve opening parameters. This reference value serves as an effective supplement or substitute for the actual measured value.

[0187] The core function of the evaluation module is to cross-validate multiple data sources to determine the reliability of the cleaning fluid flow measurement signal. When there is a significant deviation between the cleaning fluid flow measurement signal and the calculated cleaning fluid flow reference value, and the regulating valve opening parameter or pipeline pressure parameter also shows abnormalities, the evaluation module can identify situations where the measurement signal may be faulty or unreliable.

[0188] The control switching module is activated when the evaluation module determines that the cleaning fluid flow measurement signal is unreliable. At this time, the system will switch from a control mode that relies on the measurement signal to a new control mode. In this mode, the control of the cleaning fluid flow will be mainly based on the calculated reference value of the cleaning fluid flow, thereby avoiding control errors caused by sensor failure.

[0189] The alarm module promptly issues a maintenance alarm after the system detects an unreliable cleaning fluid flow measurement signal and switches control modes. This alarm notifies operators or maintenance personnel of a potential system problem, requiring inspection and maintenance to ensure the continuous and stable operation of the cleaning process.

[0190] The wine fermentation valve array cleaning fluid flow control system of this application achieves intelligent monitoring and control of the cleaning fluid flow through modular design.

[0191] The content disclosed above is only a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of protection of the present invention. Therefore, all equivalent technical changes made based on the content of the present invention specification and drawings are included within the scope of protection of the present invention. Furthermore, the elements therein can be updated as technology develops.

Claims

1. A wine fermentation valve array cleaning fluid flow control method, characterized by, The method comprises the following steps: Collecting the cleaning fluid flow measurement signal, the cleaning pump operating parameter, the regulating valve opening degree parameter and the pipeline pressure parameter; Establishing the corresponding relationship between the cleaning pump operating parameter, the regulating valve opening degree parameter and the cleaning fluid flow, and calculating the cleaning fluid flow reference value according to the cleaning pump operating parameter and the regulating valve opening degree parameter; Conducting consistency evaluation on the cleaning fluid flow measurement signal, the cleaning fluid flow reference value and the pipeline pressure parameter, and determining that the cleaning fluid flow measurement signal is not reliable when the deviation between the cleaning fluid flow measurement signal and the cleaning fluid flow reference value exceeds the preset flow deviation threshold value, and the regulating valve opening degree parameter or the pipeline pressure parameter exceeds the preset threshold value range; Switching the control mode when it is determined that the cleaning fluid flow measurement signal is not reliable, and controlling the cleaning fluid flow according to the cleaning fluid flow reference value in the switched control mode; Issuing a maintenance alarm; The consistency evaluation step comprises: Obtaining the opening degree instruction signal of the regulating valve; Carrying out high-pass filtering processing on the opening degree instruction signal of the regulating valve to obtain the rapid change component of the instruction signal; Calculating the root mean square value of the rapid change component of the instruction signal to obtain the quantization index of the rapid change component of the instruction signal; Dynamically adjusting the proportional gain and the differential gain of the controller used for consistency evaluation according to the quantization index of the rapid change component of the instruction signal; When the quantization index of the rapid change component of the instruction signal exceeds the preset upper threshold value, the proportional gain and the differential gain of the controller are reduced; When the quantization index of the rapid change component of the instruction signal continuously falls below the preset upper threshold value, the proportional gain and the differential gain of the controller are restored; Carrying out consistency evaluation on the cleaning fluid flow measurement signal, the cleaning fluid flow reference value and the pipeline pressure parameter according to the adjusted controller.

2. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 1 wherein, The step of calculating the cleaning fluid flow reference value comprises: Periodically carrying out performance self-calibration program on the cleaning pump and the regulating valve; Measuring the actual output pressure of the cleaning pump and the actual input power of the cleaning pump at the preset working condition point; Obtaining the theoretical flow value according to the cleaning pump operating parameter and the regulating valve opening degree parameter; Judging whether there is a preset allowable deviation threshold value between the actual output pressure of the cleaning pump or the actual input power of the cleaning pump and the historical reference value corresponding to the theoretical flow value; When the preset allowable deviation threshold value exists, it is determined that the cleaning pump or the regulating valve has performance attenuation; Dynamically correcting the corresponding relationship between the cleaning pump operating parameter, the regulating valve opening degree parameter and the cleaning fluid flow according to the deviation degree of the actual output pressure of the cleaning pump and the actual input power of the cleaning pump; Calculating the cleaning fluid flow reference value according to the corrected corresponding relationship between the cleaning pump operating parameter, the regulating valve opening degree parameter and the cleaning fluid flow, and according to the cleaning pump operating parameter and the regulating valve opening degree parameter.

3. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 1, wherein, Before the consistency evaluation, it further comprises identifying the running mode of the cleaning process; and adjusting the preset flow deviation threshold value and the preset threshold value range according to the identified running mode of the cleaning process.

4. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 3 wherein, The consistency evaluation further comprises: Identifying the running scene feature of the cleaning process; According to the running scene feature of the cleaning process, obtaining the dynamic fluctuation range corresponding to the running scene feature; Collecting the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter; Performing time series analysis on the collected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter, and extracting the dynamic characteristics of the collected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter; Comparing the extracted dynamic characteristics with the dynamic fluctuation range; According to the comparison result, calculating the comprehensive abnormality index; When the comprehensive abnormality index continuously exceeds the preset abnormality threshold value, determining that the cleaning liquid flow measurement signal is not reliable.

5. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 4 wherein, The step of performing time series analysis on the collected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter, and extracting the dynamic characteristics of the collected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter includes: Performing real-time monitoring on the collected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter, and identifying whether there is a short-time, instantaneous peak or burr in the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter; When the instantaneous peak or burr is identified, performing filtering processing on the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr, and eliminating or weakening the influence of the instantaneous peak or burr; During the filtering processing on the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr, suspending or adjusting the extraction of the dynamic characteristics of the affected cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter; After the filtering processing on the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr is removed, resuming the extraction of the dynamic characteristics of the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr, and performing time series analysis on the processed cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr; Performing time series analysis on the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr, and extracting the dynamic characteristics of the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter affected by the instantaneous peak or burr; Continuously performing time series analysis on the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter not affected by the instantaneous peak or burr, and extracting the dynamic characteristics of the cleaning liquid flow measurement signal, the cleaning liquid flow reference value, the regulating valve opening degree parameter and the pipeline pressure parameter not affected by the instantaneous peak or burr; Fusing all the dynamic characteristics to obtain a dynamic characteristic set reflecting the real running state of the system.

6. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 4, characterised in that, When the extracted dynamic characteristics and the dynamic fluctuation range are compared, the triggered conditions include: When the transition working condition occurs, suspending the comparison of the extracted dynamic characteristics and the dynamic fluctuation range; After the transition condition ends, the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter are stable, the dynamic characteristics and the dynamic fluctuation range of the extraction are compared.

7. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 4, characterised in that, According to the comparison result, the step of calculating the comprehensive abnormality index comprises: Identifying the deviation degree of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter in the comparison result respectively; According to the deviation degree and the relevance of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter in the historical complex fault mode, dynamically adjusting the weight of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter in the calculation of the comprehensive abnormality index; The deviation degree after the weight adjustment is aggregated to obtain the comprehensive abnormality index.

8. A wine fermentation valve array cleaning fluid flow control method as claimed in claim 7, characterised in that, The step of identifying the deviation degree comprises: Obtaining the comparison result; Normalizing the deviation degree in the comparison result to obtain a normalized deviation value; According to the normalized deviation value, identifying the deviation direction of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter; When there are multiple opposite deviation directions, calculating the weighted difference value of the normalized deviation value of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter with opposite deviation directions to obtain a counter effect amount; According to the counter effect amount, adjusting the deviation amplitude of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value, the adjusting valve opening degree parameter and the pipeline pressure parameter to obtain the adjusted deviation amplitude; The deviation direction and the adjusted deviation amplitude are taken as the deviation degree. The system comprises:

9. A wine fermentation valve array cleaning fluid flow control system for use in a wine fermentation valve array cleaning fluid flow control method as claimed in claim 1, characterised in that, The acquisition module is used to acquire the cleaning liquid flow rate measurement signal, the cleaning pump operation parameter, the adjusting valve opening degree parameter and the pipeline pressure parameter; The calculation module is used to establish the corresponding relationship between the cleaning pump operation parameter, the adjusting valve opening degree parameter and the cleaning liquid flow rate, and calculate the cleaning liquid flow rate reference value according to the cleaning pump operation parameter and the adjusting valve opening degree parameter; The evaluation module is used to evaluate the consistency of the cleaning liquid flow rate measurement signal, the cleaning liquid flow rate reference value and the pipeline pressure parameter. When the deviation between the cleaning liquid flow rate measurement signal and the cleaning liquid flow rate reference value exceeds the preset flow deviation threshold value, and the adjusting valve opening degree parameter or the pipeline pressure parameter exceeds the preset threshold range, it is determined that the cleaning liquid flow rate measurement signal is not reliable; The control switching module switches the control mode when it is determined that the cleaning liquid flow rate measurement signal is not reliable, and controls the cleaning liquid flow rate according to the cleaning liquid flow rate reference value in the switched control mode; The alarm module is used to issue a maintenance alarm; The step of evaluating the consistency comprises: Obtaining the opening instruction signal of the adjusting valve; High-pass filtering the opening instruction signal of the adjusting valve to obtain the fast change component of the instruction signal; Calculating the root mean square value of the fast change component of the instruction signal to obtain a quantization index of the fast change component of the instruction signal; According to the quantization index of the fast change component of the instruction signal, dynamically adjusting the proportional gain and the differential gain of the controller used for consistency evaluation; ​ When the quantification index of the fast change component of the instruction signal exceeds a preset upper threshold, the proportional gain and the differential gain of the controller are reduced; When the quantification index of the fast change component of the instruction signal continuously falls below the preset upper threshold, the proportional gain and the differential gain of the controller are restored; According to the adjusted controller, the consistency of the cleaning liquid flow measurement signal, the cleaning liquid flow reference value and the pipeline pressure parameter is evaluated.

Images