Anti-blocking measurement and control system for wedge-shaped flow meter

Abstract

The invention discloses an anti-blocking measurement and control system for a wedge-shaped flow meter, particularly relates to the technical field of industrial flow monitoring, and aims to solve the problems of flow measurement distortion and maintenance lag caused by winding type local blockage due to the fact that flexible impurities such as fibers and viscous substances in fluid cannot be effectively identified in the prior art. According to the system, through frequency domain energy analysis and time domain multi-scale entropy attenuation fitting of differential pressure signals, the low-frequency energy proportion and high-frequency noise abrupt change characteristics of flow field distortion are extracted, a winding risk coefficient is generated through dynamic fusion in combination with fluid characteristics, directional pulse cleaning is accurately triggered, and a signal zero point is calibrated in a closed loop mode. According to the technical scheme, signal detection, feature analysis, risk judgment and control execution full links are covered, real blockage and instantaneous interference are effectively distinguished, the limitation of traditional threshold value alarm or periodic cleaning is avoided, and the measurement precision and maintenance efficiency under the complex working condition are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial flow monitoring, and more particularly to an anti-clogging measurement and control system for a wedge-shaped flowmeter. Background Art

[0002] Wedge flowmeters are widely used in industrial flow monitoring scenarios of fluids containing impurities due to their simple structure and wear resistance, such as sewage treatment, papermaking, food processing and other fields. In the prior art, conventional means to address the problem of fluid impurity blockage often use mechanical filtration, regular cleaning or the addition of protective covers to reduce the direct impact or deposition of solid particles on the measuring element. However, the above methods are difficult to effectively deal with fibers, viscous substances or flexible impurities (such as pulp fibers, hair, etc.) present in the fluid. Such impurities are easy to adhere to the edge of the wedge structure or the surface of the flow channel, and form entanglement and adhesion with the flow of the fluid, resulting in local deformation of the cross-sectional area of ​​the flow channel, but it has not yet reached the critical state of complete blockage.

[0003] The existing anti-clogging technology has a problem: when fibers or soft impurities in the fluid form a tangled local blockage in the wedge-shaped area, the system cannot identify the abnormality in time because the degree of flow channel deformation does not trigger the preset pressure difference alarm threshold. At the same time, such local entanglement will change the flow field distribution characteristics, causing the flow measurement value to continue to deviate from the true value. Traditional control logic based on fixed thresholds or periodic cleaning has difficulty distinguishing the cause of the signal abnormality, ultimately leading to accumulated measurement errors and delayed maintenance. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an anti-clogging measurement and control system for a wedge-shaped flowmeter to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An anti-clogging measurement and control system for a wedge-shaped flowmeter, comprising:

[0007] A pressure difference detection module is used to detect the pressure difference signal between the inlet and outlet of the wedge flowmeter and generate a corresponding frequency spectrum distribution diagram;

[0008] Feature extraction module, used to extract the low-frequency energy ratio of flow field distortion and the number of high-frequency noise mutation points from the spectrum distribution diagram;

[0009] A trigger determination module is used to determine whether to trigger winding type congestion analysis based on whether the proportion of low-frequency energy exceeds a first threshold and the number of high-frequency noise mutation points is less than a second threshold;

[0010] The entropy analysis module is used to perform multi-scale sample entropy analysis on the pressure difference signal when the winding type blockage analysis is triggered, fit the entropy attenuation slope of each scale, and generate the complexity attenuation rate;

[0011] The risk fusion module is used to generate the entanglement risk coefficient based on the complexity decay rate and the proportion of low-frequency energy;

[0012] The control execution module is used to control the cleaning device to perform directional pulse flushing on the wedge-shaped area when the entanglement risk coefficient exceeds a third threshold value, and to recalibrate the zero point of the pressure difference signal after flushing.

[0013] In a preferred embodiment, detecting the pressure difference signal between the inlet and outlet of the wedge flowmeter and generating a corresponding frequency spectrum distribution diagram includes:

[0014] The real-time differential pressure signal between the inlet and outlet of the wedge flowmeter is continuously collected through the pressure sensor;

[0015] preprocessing the pressure difference signal to remove high-frequency noise and baseline drift, thereby generating a preprocessed pressure difference signal;

[0016] Based on the preprocessed pressure difference signal, a frequency spectrum distribution diagram covering a preset frequency band is generated through fast Fourier transform.

[0017] In a preferred embodiment, the low-frequency energy ratio of the flow field distortion and the number of high-frequency noise mutation points are extracted from the spectrum distribution diagram, including:

[0018] Integrate and calculate the normalized amplitude energy within a preset low-frequency band of the spectrum distribution diagram to obtain a low-frequency energy value;

[0019] Integrate and calculate the normalized amplitude energy within a preset high frequency band of the spectrum distribution diagram to obtain a high frequency energy value;

[0020] The ratio of the low-frequency energy value to the sum of the low-frequency energy value and the high-frequency energy value is taken as the low-frequency energy proportion;

[0021] The number of amplitude energy differences between adjacent frequency points in the high-frequency band exceeding the preset mutation threshold is detected and counted as the number of high-frequency noise mutation points.

[0022] In a preferred embodiment, determining whether to trigger winding type blockage analysis is based on whether the low-frequency energy ratio exceeds a first threshold and the number of high-frequency noise mutation points is less than a second threshold, including:

[0023] Compare the low-frequency energy ratio with a preset first threshold, and compare the number of high-frequency noise mutation points with a preset second threshold;

[0024] When the proportion of low-frequency energy exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold, the winding type congestion analysis is triggered, otherwise it is not triggered.

[0025] In a preferred embodiment, the first threshold is dynamically adjusted according to a preset multiple range of the mean low-frequency energy proportion under historical normal operating conditions corresponding to the fluid type; the second threshold is set according to the high quantile point of the statistical distribution of the number of high-frequency noise mutation points under non-blocking conditions.

[0026] In a preferred embodiment, when the winding type blockage analysis is triggered, a multi-scale sample entropy analysis is performed on the pressure difference signal, and the attenuation slope of the entropy value of each scale is fitted to generate a complexity attenuation rate, including:

[0027] Based on the dynamic fluctuation characteristics of the pressure difference signal, the coarse-graining time scale range is divided according to the fluid turbulence intensity;

[0028] Generate the corresponding coarse-grained sequence according to the coarse-grained time scale range;

[0029] Calculate the sample entropy value for each coarse-grained sequence and detect its non-stationary decay trend with increasing scale;

[0030] In view of the non-stationary attenuation trend, piecewise linear regression is used to fit the attenuation slopes of the initial segment and the stable segment respectively; the absolute value of the difference between the attenuation slopes of the initial segment and the stable segment is taken as the complexity attenuation rate.

[0031] In a preferred embodiment, low turbulence intensity corresponds to long-scale sequences, and high turbulence intensity corresponds to short-scale sequences.

[0032] In a preferred embodiment, the entanglement risk coefficient is generated based on the complexity decay rate and the low-frequency energy ratio, including:

[0033] Dynamically allocate the fusion weight of complexity attenuation rate and low-frequency energy proportion according to the viscosity characteristics corresponding to the fluid type;

[0034] The complexity attenuation rate and the low-frequency energy ratio are normalized to eliminate the dimensional differences.

[0035] The normalized complexity decay rate and the low-frequency energy ratio are linearly weighted and summed according to the assigned weights to generate the entanglement risk coefficient;

[0036] When the sum of the fusion weights exceeds the preset range, they are redistributed according to the weight ratio until the sum reaches the set value.

[0037] In a preferred embodiment, the weight of the complexity decay rate corresponding to high viscosity fluid is increased.

[0038] In a preferred embodiment, when the entanglement risk coefficient exceeds a third threshold, controlling the cleaning device to perform directional pulse flushing on the wedge-shaped area and recalibrating the zero point of the pressure difference signal after flushing includes:

[0039] When the entanglement risk coefficient exceeds a third threshold, the intensity and duration of the flushing pulse are dynamically adjusted according to a mapping relationship between the fluid type and the entanglement risk coefficient;

[0040] After the flushing is completed, a preset time window is delayed to wait for the fluid flow to stabilize, and then the differential pressure signal in the stable state is collected to calculate the zero offset;

[0041] If the zero offset exceeds the preset tolerance range, repeat the flushing and calibration process until the offset meets the standard;

[0042] After calibration is completed, the zero point reference value in the historical database is updated for subsequent detection cycles.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention effectively solves the problem that traditional anti-blocking technology has difficulty in identifying entanglement-type local blockages and misjudging signal anomalies through multi-dimensional signal feature extraction and dynamic fusion mechanism. The system adopts dual analysis of frequency domain energy distribution and time domain complexity attenuation, combined with a weight distribution strategy that is adaptive to fluid characteristics, to significantly improve detection sensitivity and adaptability to working conditions. The energy concentration effect of flow field distortion is captured by low-frequency energy proportion, particle interference is filtered by high-frequency noise mutation points, and the dynamic instability rate is quantified by multi-scale sample entropy, so that the system can distinguish the difference characteristics between fiber entanglement and conventional impurity scouring, and avoid false triggering caused by a single threshold or fixed cleaning logic. At the same time, the closed-loop calibration mechanism automatically corrects the zero point of the pressure difference signal after cleaning, eliminates the residual error of the flow field disturbance, and ensures that the measurement data is continuously reliable.

[0045] 2. Dynamically adjusted cleaning pulse parameters adapt to different fluid viscosities and blockage levels, preventing both error accumulation due to insufficient cleaning and energy waste from over-flushing. Through the nonlinear integration of entanglement risk factors, the system triggers precise cleaning at the early stages of flow field distortion, resolving the maintenance delays associated with traditional methods due to delayed flow path deformation. This significantly improves the long-term stability and measurement accuracy of the wedge flowmeter in complex impurity environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The figure is a structural schematic diagram of an anti-clogging measurement and control system for a wedge-shaped flowmeter according to the present invention. DETAILED DESCRIPTION

[0047] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0048] Example: Figure 1 The present invention provides a structural schematic diagram of an anti-clogging measurement and control system for a wedge-shaped flowmeter, which includes the following modules:

[0049] A pressure difference detection module is used to detect the pressure difference signal between the inlet and outlet of the wedge flowmeter and generate a corresponding frequency spectrum distribution diagram;

[0050] Feature extraction module, used to extract the low-frequency energy ratio of flow field distortion and the number of high-frequency noise mutation points from the spectrum distribution diagram;

[0051] A trigger determination module is used to determine whether to trigger winding type congestion analysis based on whether the proportion of low-frequency energy exceeds a first threshold and the number of high-frequency noise mutation points is less than a second threshold;

[0052] The entropy analysis module is used to perform multi-scale sample entropy analysis on the pressure difference signal when the winding type blockage analysis is triggered, fit the entropy attenuation slope of each scale, and generate the complexity attenuation rate;

[0053] The risk fusion module is used to generate the entanglement risk coefficient based on the complexity decay rate and the proportion of low-frequency energy;

[0054] The control execution module is used to control the cleaning device to perform directional pulse flushing on the wedge-shaped area when the entanglement risk coefficient exceeds a third threshold value, and to recalibrate the zero point of the pressure difference signal after flushing.

[0055] The differential pressure detection module uses a sensor to collect the differential pressure signals between the inlet and outlet in real time. After preprocessing, it generates a spectrum distribution map and transmits the spectrum data to the feature extraction module. The feature extraction module extracts the low-frequency energy percentage and the number of high-frequency noise mutation points that indicate flow field distortion from the spectrum distribution map and outputs these to the trigger determination module. The trigger determination module determines whether to initiate the subsequent analysis process based on the dual conditions of whether the low-frequency energy percentage exceeds a first threshold and whether the number of high-frequency noise mutation points is less than a second threshold. If a trigger is determined, the entropy analysis module performs multi-scale sample entropy analysis on the raw differential pressure signal, fitting the entropy decay slopes at each scale to generate a complexity decay rate. The result is input into the risk fusion module. The risk fusion module dynamically assigns weights based on fluid type and performs a nonlinear fusion of the normalized complexity decay rate and the low-frequency energy percentage to generate an entanglement risk coefficient. The control execution module monitors the risk coefficient in real time. When it exceeds a third threshold, it controls the cleaning device to perform a directed flush based on the pulse parameters mapped to the fluid type. After flushing, it delays for a specific time window to stabilize the flow field. The pressure differential signal is then re-collected to calibrate the zero point, completing closed-loop control.

[0056] Each module forms a collaborative mechanism through the series connection of data streams and logic chains: differential pressure detection and feature extraction enable multi-dimensional signal analysis, trigger judgment avoids the risk of misjudgment of a single parameter, entropy analysis captures the characteristics of dynamic instability in the time domain, risk fusion enhances the robustness of complex working conditions, and control execution combines dynamic parameters with closed-loop calibration to address the defects of overload or insufficient cleaning in traditional methods. This connection relationship breaks through the limitations of conventional threshold alarms or fixed cleaning strategies through joint frequency-time domain analysis, multi-threshold collaborative judgment, and fluid-adaptive dynamic control. In particular, it addresses the hidden, slow-changing, and fluid-dependent nature of entanglement-type blockages, achieving full-link closed-loop management from signal perception to precise execution.

[0057] Detect the pressure difference signal between the inlet and outlet of the wedge flowmeter and generate the corresponding spectrum distribution diagram. The specific implementation is as follows:

[0058] When a pressure sensor continuously collects the real-time differential pressure signal between the inlet and outlet of a wedge flowmeter, the specific implementation method is as follows: a differential pressure sensor is installed on the outer wall of the pipe at the inlet and outlet of the wedge flowmeter, with the installation spacing between the two sensors being 1 to 3 times the pipe diameter to avoid interference from fluid disturbances. The sampling frequency of the pressure sensor is set to 1000 Hz to 5000 Hz, covering the main frequency range of fluid flow velocity fluctuations. During the acquisition process, the pressure sensor outputs a continuous voltage analog signal, which is converted into a digital signal by an analog-to-digital converter with a conversion accuracy of 12 to 16 bits, ensuring that the dynamic range of the differential pressure signal meets the measurement requirements.

[0059] When the pressure difference signal is preprocessed to remove high-frequency noise and baseline drift, the specific implementation method is as follows: first, the original pressure difference signal is filtered by a Butterworth low-pass filter, and the cutoff frequency of the filter is set to one-tenth to one-fifth of the sampling frequency. For example, when the sampling frequency is 2000 Hz, the cutoff frequency is 200 Hz to 400 Hz to filter out the sensor circuit noise and high-frequency fluid turbulence noise. Secondly, the filtered signal is corrected for baseline drift. The baseline drift correction method is: intercept the signal segment of the fluid static or stable flow period in the pressure difference signal as a benchmark, calculate the mean of the segment as the baseline offset, and subtract the offset from the entire signal. If the fluid cannot be completely still, the sliding window polynomial fitting method is used, with a window length of 10 seconds to 30 seconds, and a quadratic polynomial fitting is performed on the signal in the window. The low-frequency trend of the fitting curve is deducted as the baseline drift to generate a preprocessed pressure difference signal.

[0060] When generating a spectrum distribution plot covering a preset frequency band using a fast Fourier transform based on a preprocessed pressure differential signal, the specific implementation method is as follows: the preprocessed pressure differential signal is divided into multiple consecutive time windows according to the time series, with each time window length ranging from 1 to 5 seconds and an overlap ratio of 20% to 50% between adjacent windows. A Hanning window function is applied to the signal within each time window to reduce spectral leakage. The windowed signal is then subjected to a fast Fourier transform to calculate its amplitude spectrum. The preset frequency band is set based on the fluid type and pipe size. For example, for a water-based fluid with a pipe diameter of 50 mm to 200 mm, the preset frequency band covers 0 Hz to 100 Hz, where 0 Hz to 10 Hz is defined as the low frequency band and 10 Hz to 100 Hz is defined as the high frequency band. The horizontal axis of the spectrum distribution plot is the frequency value, and the vertical axis is the normalized amplitude energy. The amplitude energy value at each frequency point is the square of the modulus of the Fourier transform result at that frequency divided by the signal length.

[0061] During the preprocessing process, the order of the Butterworth low-pass filter is set to 4th to 8th order to achieve a steep transition between the passband and the stopband and avoid signal distortion. For example, when the cutoff frequency is 300 Hz, a 6th-order Butterworth filter is used, and its passband fluctuation is less than 0.1 dB and the stopband attenuation is greater than 60 dB. The sliding window polynomial fitting method used in baseline drift correction has its polynomial order adjusted according to the stability of the fluid flow: for stable flow conditions, 1st to 2nd order polynomials are used; for fluctuating flow conditions, 3rd to 4th order polynomials are used. If there is still a residual offset in the signal after baseline drift correction, it is further processed by mean zeroing, that is, the mean of the entire signal sequence is calculated and subtracted.

[0062] The specific parameters of the fast Fourier transform are set as follows: the number of signal points in each time window is determined by the sampling frequency and window length. For example, when the sampling frequency is 2000 Hz and the window length is 2 seconds, the number of signal points is 4000 points. The number of fast Fourier transform points is expanded to 4096 points to match the integer power of 2. The frequency resolution of the spectrum distribution diagram is the sampling frequency divided by the number of transformation points. For example, when the sampling frequency is 2000 Hz and the number of transformation points is 4096, the frequency resolution is 0.488 Hz. The spectrum energy is normalized by dividing the amplitude value of each frequency point by the maximum possible amplitude value of the signal. The maximum possible amplitude value is determined by the range of the analog-to-digital converter. For example, when the analog-to-digital converter range is ±5 volts, the maximum possible amplitude value is 5 volts.

[0063] After generating the spectrum distribution diagram, the system automatically records the spectrum data of each time window and stores it in matrix form in chronological order. The rows of the matrix correspond to the time window index, the columns correspond to the frequency point index, and the matrix element values ​​are normalized amplitude energy. The division of the preset frequency bands is based on the fluid flow characteristics and the common interference frequency range. For example, in the sewage treatment scenario, the upper limit of the high frequency band of the preset frequency band is set to 80 Hz to avoid the influence of water pump vibration noise. If the fluid type or pipe size changes in actual application, the preset frequency band can be manually adjusted through the user interface, and the adjusted frequency band data is saved to the system configuration file for subsequent analysis and call.

[0064] Extract the low-frequency energy ratio of flow field distortion and the number of high-frequency noise mutation points from the spectrum distribution diagram. The specific implementation is as follows:

[0065] When extracting the proportion of low-frequency energy of flow field distortion from the spectrum distribution diagram, the specific implementation method is as follows: the normalized amplitude energy is integrated and calculated within a preset low-frequency band. The frequency range of the preset low-frequency band is dynamically adjusted according to the fluid type. For example, it is set to 0Hz to 10Hz for water-based fluids and 0Hz to 5Hz for oil-based high-viscosity fluids. The integral calculation adopts the trapezoidal integration method to adapt to the uneven frequency intervals in the spectrum distribution diagram. The specific steps are as follows: the normalized amplitude energy values ​​of two adjacent frequency points are added and multiplied by the frequency interval, and then divided by 2, and the low-frequency energy value is accumulated point by point. The frequency interval is determined by the spectral resolution of the fast Fourier transform. For example, when the spectral resolution is 0.488Hz, there are 20 intervals in the range of 0Hz to 10Hz, and the integral contribution value of each interval is (energy value 1 + energy value 2) × 0.488 / 2. The source of the normalized amplitude energy value is the pre-processed differential pressure signal generated by fast Fourier transform. Its value is determined by the ratio of the 16-bit analog-to-digital converter range ±5V to the digital range of 32767, that is, 1 corresponds to 5V.

[0066] To extract the number of high-frequency noise mutation points, the specific implementation involves detecting the number of adjacent frequency points within a preset high-frequency band whose amplitude energy differences exceed a preset mutation threshold. The preset high-frequency band is divided to cover the primary noise interference frequency band. For example, in industrial wastewater containing sand, the range is set to 20Hz to 150Hz to avoid interference from the pump fundamental frequency. The difference between adjacent frequency points is calculated using the absolute value method: |energy value n + 1 - energy value n|, where n is the frequency point index. The preset mutation threshold can be set in two modes: static and dynamic. In the static mode, a fixed value is preset based on the type of fluid impurity, such as 0.03 for fiber impurities and 0.08 for particulate impurities. In the dynamic mode, 10% of the maximum energy value within the high-frequency band is used as an adaptive threshold. For example, if the maximum energy value in the high-frequency band is 0.5, the dynamic threshold is 0.05. During the detection process, if a difference exceeds the threshold, a counter is incremented by 1, and the final statistical result is the number of mutation points.

[0067] When calculating the proportion of low-frequency energy, the specific implementation method is: convert the ratio of the low-frequency energy value to the total energy value (low-frequency energy value + high-frequency energy value) into a percentage. The calculation method of the high-frequency energy value is the same as the low-frequency energy value, and the integration interval is the preset high-frequency band. For example, in the pulp pipeline of a paper mill, the integral result of the low-frequency energy value is 0.8, and the high-frequency energy value is 1.2, then the low-frequency energy proportion is 0.8 / (0.8+1.2)×100%=40%. The calculation of the total energy value excludes interference components outside the preset frequency band, such as ultrasonic noise above 150Hz. If the total energy value is zero (not feasible in theory, but may be caused by signal loss), the proportion value is forced to zero and triggers the sensor fault alarm.

[0068] The dynamic adjustment of the preset frequency band is achieved through the user interface. After the user enters the fluid type (such as "sewage" or "crude oil"), the system automatically matches the built-in frequency band parameter library. For example, when "sewage" is selected, the 0-10Hz low frequency band and the 10-100Hz high frequency band are loaded; when "crude oil" is selected, the 0-5Hz low frequency band and the 5-50Hz high frequency band are loaded. The frequency band parameter library is established based on multi-condition experimental data. The experimental method includes: collecting pressure difference signals under clean flow field and simulated blocked flow field respectively, comparing the difference in spectral energy distribution, and determining the energy concentration frequency band. For example, fiber winding causes the 0-10Hz energy to increase by more than 30%, while particle accumulation causes the number of 10-100Hz mutation points to increase by 50%.

[0069] When detecting high-frequency noise mutation points, the fault tolerance mechanism for transient signal anomalies (such as sensor disconnection) is as follows: if the number of mutation points in a single time window exceeds 50% of the total number of frequency points in the high-frequency band, the signal is considered abnormal, the current window data is discarded, and the previous window data is interpolated and replaced. The interpolation method is linear extrapolation. For example, if the previous window has 20 mutation points and the current window detects 100 mutation points due to an anomaly, the number of replacement points is 20.

[0070] The hardware requirements for integral calculation and mutation point detection include a processor supporting floating-point operations (such as an ARM Cortex-M7) and at least 100KB of memory for storing the spectrum matrix data. Software requirements include open-source math libraries (such as the trapz function in NumPy 1.21.5). During the calculation process, if the processor load exceeds 80%, the time window overlap ratio of the spectrum distribution graph is automatically reduced to 20% to ensure real-time performance.

[0071] The optimized mutation threshold is validated through cross-validation. 100 sets of historical data (50 under normal operating conditions and 50 under blocked conditions) are collected. The false alarm and false negative rates at different thresholds are tested, and the threshold that yields the highest F1 score is selected. For example, in the fiber impurity scenario, a threshold of 0.04 achieves an F1 score of 0.92 (95% precision and 89% recall). The threshold data is stored in non-volatile memory and automatically loaded upon system startup.

[0072] Whether to trigger winding type congestion analysis is determined based on whether the low-frequency energy ratio exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold. The specific implementation is as follows:

[0073] When determining whether to trigger the winding type blockage analysis based on whether the low-frequency energy ratio exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold, the specific implementation method is as follows: the first threshold is dynamically adjusted according to the preset multiple range of the average low-frequency energy ratio under historical normal operating conditions corresponding to the fluid type. The historical normal operating condition data is obtained by long-term monitoring of the flow signal in a non-blocked state and stored in an embedded database (such as SQLite). The data retention period is 3 months to 1 year. The preset multiple range is set according to the fluid viscosity characteristics: for low-viscosity fluids (such as water-based solutions), the multiple range is 1.2 times to 1.5 times; for high-viscosity fluids (such as crude oil), the multiple range is 1.1 times to 1.3 times. For example, when the average low-frequency energy ratio under historical normal conditions is 30%, the first threshold of the low-viscosity fluid is set to 36% to 45%. The adjusted first threshold is written to the system configuration file and automatically loaded when the fluid type is switched.

[0074] The second threshold is set based on the high percentile of the statistical distribution of the number of high-frequency noise mutation points under non-congestion conditions. The high percentile is determined by statistically analyzing the distribution of the number of mutation points during non-congestion periods in historical data. The specific method is: collect at least 100 sets of non-congestion data, arrange them in ascending order, and take the 90th percentile value as the second threshold. For example, if the number of 90th mutation points in 100 sets of data is 50 times / second, the second threshold is set to 50 times / second. If there are less than 100 sets of historical data, the high percentile is estimated by linear interpolation: when the amount of data is N sets, the 90th percentile position is 0.9×(N+1), and the weighted average of the two adjacent data at this position is taken.

[0075] The comparison logic for the low-frequency energy percentage and the number of high-frequency noise mutation points is as follows: the real-time calculated low-frequency energy percentage is compared with the dynamically adjusted first threshold, and the real-time number of high-frequency noise mutation points is compared with the second threshold. If and only if the low-frequency energy percentage exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold, it is determined to meet the characteristics of a winding-type congestion, triggering subsequent analysis. For example, if the real-time low-frequency energy percentage is 42% (the first threshold is 40%) and the real-time number of high-frequency mutation points is 30 times / second (the second threshold is 50 times / second), analysis is triggered if both conditions are met.

[0076] The exception handling mechanism for dynamic threshold adjustment includes the following: If there are fewer than 10 sets of historical normal operating condition data, the first threshold is set to the default multiple of 1.3, and an insufficient data alarm is issued. If the historical data for the number of high-frequency noise mutation points is all zero (theoretically impossible, possibly due to sensor failure), a preset empirical value (such as 40 times / second) is temporarily used as the second threshold, and the sensor calibration procedure is initiated. The threshold update cycle is set to weekly or monthly, and is adjusted based on the stability of the system operating environment: weekly updates in high-pollution scenarios such as chemical industry, and monthly updates in stable scenarios such as water treatment.

[0077] The hardware requirements for the decision logic include a microprocessor with floating-point support (such as an ARM Cortex-M4) and at least 50KB of memory for storing threshold parameters and historical data indexes. Software requirements include open-source statistical libraries (such as the percentile function in SciPy 1.7.3) and a linear interpolation algorithm. During the decision process, if the real-time data is abnormal (such as the low-frequency energy percentage exceeding 100% or the number of high-frequency mutation points being negative), the current data is discarded and re-collected. Three consecutive abnormalities trigger a system self-check.

[0078] When the winding blockage analysis is triggered, a multi-scale sample entropy analysis is performed on the pressure difference signal, and the attenuation slope of the entropy value at each scale is fitted to generate the complexity attenuation rate. The specific implementation is as follows:

[0079] When the winding type blockage analysis is triggered, a multi-scale sample entropy analysis is performed on the pressure difference signal, and the attenuation slope of the entropy value of each scale is fitted to generate the specific implementation method of the complexity attenuation rate: first, the fluid turbulence intensity is calculated based on the dynamic fluctuation characteristics of the pressure difference signal. The fluid turbulence intensity is quantified by the variance value of the pressure difference signal within a preset time window. The larger the variance value, the higher the turbulence intensity. For example, when the variance value of the pressure difference signal is in the range of 0.1 to 0.5, it is judged as low turbulence intensity, and when the variance value is in the range of 0.5 to 2.0, it is judged as high turbulence intensity. The coarse-grained time scale range is divided according to the turbulence intensity. The time window length of the long-scale sequence corresponding to low turbulence intensity is 2 seconds to 5 seconds, and the time window length of the short-scale sequence corresponding to high turbulence intensity is 0.5 seconds to 1 second. The time window length for turbulence intensity calculation is fixed at 10 seconds, which is independent of the window of the subsequent coarse-grained step.

[0080] The coarse-grained sequence is generated by segmenting the original pressure difference signal into non-overlapping time window segments, then averaging the data points within each segment to generate the coarse-grained sequence. For example, when the time window length is 2 seconds and the sampling frequency is 1000 Hz, each segment contains 2000 data points, and the coarse-grained sequence is the mean of each 2000 data points. Long-scale sequences with low turbulence intensity retain low-frequency trends, while short-scale sequences with high turbulence intensity retain high-frequency details. The standard deviation of the coarse-grained sequence is calculated as an indicator of the degree of dispersion of each data segment and is used for subsequent adjustment of the sample entropy parameter.

[0081] When calculating the sample entropy for each coarse-grained sequence, the calculation parameters for the sample entropy are set as follows: embedding dimension m = 2 (according to the general standard for nonlinear time series analysis) and tolerance r = 0.2 times the standard deviation of the coarse-grained sequence. The sample entropy reflects the complexity of the sequence, with lower values ​​indicating a more regular sequence. For example, the sample entropy of a long-scale coarse-grained sequence might be 1.2, while the sample entropy of a short-scale sequence might be 1.8. When detecting the non-stationary decay trend of sample entropy with increasing scale, non-stationarity is determined by the amplitude of the entropy fluctuation with scale: if the difference in entropy between adjacent scales exceeds 1.5 times the average difference between the previous three scales, it is considered a non-stationary decay. This judgment rule has been validated with historical data and has an accuracy rate exceeding 90%.

[0082] The calculation method of sample entropy is:

[0083]

[0084] Where: a represents the coarse-grained sequence X = [x1, x2, ..., x N ], containing N data points; b = m represents the embedding dimension (typical value is 2), which is used to define the vector length; c = r represents the tolerance parameter, which is 0.2 times the standard deviation of the coarse-grained sequence; E m(d) represents the proportion of all vector pairs whose distance is less than r in dimension m.

[0085] The specific calculation steps are as follows:

[0086] Construct a vector of dimension m:

[0087]

[0088] Among them, Represents a vector consisting of m consecutive data starting from the i-th point in the sequence x;

[0089] Probability calculation:

[0090]

[0091] Among them, E m (r) represents the average similarity ratio of all vector pairs under dimension m; Represents the similarity ratio of the i-th vector to all other vectors in dimension m; similarly, E is calculated m+1 (r) The offspring enters the main formula to obtain the sample entropy value; E m+1 (r) represents the average similarity ratio of all vector pairs in dimension m+1, which is calculated in the same way as E m (r) Same.

[0092] When using piecewise linear regression to fit the decay slopes of the initial and stable segments for non-stationary decay trends, the segmentation method is as follows: the sample entropy value sequence for the first one-third of the scale is divided into the initial segment, and the last two-thirds of the scale is divided into the stable segment. For example, when the total number of scales is six, the initial segment includes the first two scales, and the stable segment includes the last four scales. The linear regression fitting objective is to establish a linear relationship between the scale index and the sample entropy value. The slope of the initial segment reflects the rate of rapid complexity decline, while the slope of the stable segment reflects the rate of sustained decline. The absolute value of the difference in decay slopes is calculated as the absolute value of the slope of the initial segment minus the slope of the stable segment. For example, if the slope of the initial segment is -0.3 and the slope of the stable segment is -0.1, the difference is 0.2, and the absolute value is the complexity decay rate.

[0093] The exception handling mechanism includes the following: If the standard deviation of the coarse-grained sequence is zero (theoretically unfeasible, possibly due to sensor failure), the sample entropy calculation for the current sequence is skipped and the data is marked as invalid. If the sample entropy value shows an increasing trend with increasing scale (contradicting the theoretical attenuation), it is determined to be a data anomaly and triggers re-collection. Parameter adjustment is based on the following: the turbulence intensity threshold is determined by the variance distribution statistics of normal and blocked conditions in historical data. For example, the variance of normal conditions is concentrated between 0.1 and 0.8, while the variance of blocked conditions is concentrated between 0.8 and 2.0. Threshold data is stored in an embedded database and supports dynamic loading by fluid type.

[0094] Hardware requirements include a multi-threaded microcontroller (e.g., STM32H743, 400MHz, 512KB of memory) for parallel computation of sample entropy values ​​at different scales. Software requirements include an open-source entropy library (e.g., EntropyLibrary 2.3). During the computation, if the microcontroller's memory usage exceeds 80%, the number of coarse-grained scales is automatically reduced to a preset minimum (e.g., 4) to ensure real-time performance.

[0095] The specific division rules for the coarse-grained time scale range are adaptively adjusted based on the fluid type: for water-based fluids, the turbulence intensity threshold is set at a variance of 0.5. That is, when the variance is below 0.5, the long scale (2 to 5 seconds) is used, and when it is above 0.5, the short scale (0.5 to 1 second) is used. For oil-based fluids, the threshold is set at a variance of 0.3, as high-viscosity fluids have lower turbulence intensity overall. For example, when the variance is 0.4 for oil-based fluids, the long scale analysis is still used to capture slow flow distortions. The adaptive adjustment logic is implemented in a configuration file, and the user selects the fluid type through the human-machine interface.

[0096] The tolerance parameter r in the sample entropy calculation is adjusted based on the dynamic range of the coarse-grained sequence: when the sequence standard deviation is less than 0.1, r = 0.15; when the standard deviation is greater than 0.1, r = 0.2. This rule prevents entropy distortion caused by excessively large tolerances for low-volatility sequences. For example, when the coarse-grained sequence standard deviation is 0.08, r = 0.15; when the standard deviation is 0.12, r = 0.2. The adjusted r value is written to a temporary cache for subsequent calculations.

[0097] The ratio of the initial segment to the stable segment in piecewise linear regression is optimized based on historical data. Analysis of 100 sets of normal and congested data revealed that the rate of entropy decrease in the initial phase of congestion (the initial segment) is two to three times that of the stable segment. Therefore, the initial segment is set to one-third of the total scale. For a total of nine scales, the initial segment is the first three, and the stable segment is the last six. The ratio parameters are stored in read-only memory to prevent runtime tampering.

[0098] The application logic of the complexity decay rate is as follows: when the decay rate exceeds a dynamic threshold (e.g., 0.25), it is determined to be a high-risk entanglement blockage. The dynamic threshold is set based on the fluid type and pipe diameter. For example, the threshold for oil-based fluids in a DN100 pipe is 0.2, and the threshold for sewage fluids in a DN200 pipe is 0.3. Threshold data is stored in non-volatile memory and supports calibration by field engineers through the human-machine interface. During calibration, the system automatically records the last 30 days of historical decay rate data and generates recommended threshold values ​​for engineers to confirm.

[0099] Boundary condition handling includes: If the differential pressure signal variance exceeds the sensor range (e.g., variance > 5.0), it is determined to be a hardware failure, the multi-scale analysis function is disabled, and an alarm is issued. If the sample entropy value cannot be calculated due to data loss, the entropy value of the previous period is interpolated. The interpolation method is linear extrapolation. For example, if the entropy value of the previous period is 1.5 and the current period is missing, the interpolated value is 1.5. The interpolated data is marked as estimated for reference in subsequent steps.

[0100] Based on the complexity attenuation rate and the proportion of low-frequency energy, the entanglement risk coefficient is generated. The specific implementation is as follows:

[0101] The complexity decay rate and low-frequency energy proportion are weighted and fused to generate the entanglement risk coefficient. Specifically, weights are dynamically assigned based on the viscosity characteristics of the fluid type. Fluid types are obtained through viscosity sensors or a preset fluid parameter library. Viscosity characteristics are categorized as low viscosity (less than 100 cP), medium viscosity (100 cP to 1000 cP), and high viscosity (greater than 1000 cP). The weighting rule is as follows: for low-viscosity fluids, the low-frequency energy proportion is weighted 60% to 70%, and the complexity decay rate is weighted 30% to 40%. For high-viscosity fluids, the low-frequency energy proportion is weighted 30% to 40%, and the complexity decay rate is weighted 60% to 70%. For example, when the fluid is detected as high-viscosity crude oil (viscosity 1200 cP), the complexity decay rate is weighted 65%, and the low-frequency energy proportion is weighted 35%. Weight data is stored in an embedded database and can be dynamically loaded based on real-time viscosity test results.

[0102] The complexity decay rate and low-frequency energy percentage are normalized by calculating their maximum and minimum values ​​under the current operating conditions, respectively, and mapping the real-time values ​​to a range of 0 to 1. For example, if the complexity decay rate has a maximum value of 0.8, a minimum value of 0.1, and a current value of 0.5 in historical data, the normalized result is (0.5-0.1) / (0.8-0.1) = 0.57. If the low-frequency energy percentage has a maximum value of 50%, a minimum value of 10%, and a current value of 30%, the normalized result is (30-10) / (50-10) = 0.5. The normalization parameters are updated every 24 hours, based on the extreme values ​​from the last seven days of historical data.

[0103] When the normalized complexity attenuation rate and the low-frequency energy ratio are linearly weighted and summed according to the assigned weights, the calculation formula is:

[0104] crx=fv*Az+sf*Bz

[0105] Among them, crx is the entanglement risk coefficient, fv is the normalized complexity decay rate, sf is the normalized low-frequency energy ratio, Az is the weight A, and Bz is the weight B.

[0106] The sum of weights A and B is forced to be 100%. For example, if weight A is 65%, weight B is 35%, the normalized complexity decay rate is 0.57, and the normalized low-frequency energy fraction is 0.5, then the risk factor = 0.57 × 0.65 + 0.5 × 0.35 = 0.555. If the input weights do not sum to 100% due to a configuration error (for example, weight A = 70%, weight B = 40%), the system automatically scales them proportionally: weight A = 70 / (70+40) = 63.6%, weight B = 40 / (70+40) = 36.4%, ensuring the total is 100%.

[0107] The basis for adjusting the dynamic weight distribution is: through experimental testing of the sensitivity of various parameters under different viscosity fluids, the proportion of low-frequency energy in low-viscosity fluids is more sensitive to blockage, and the complexity attenuation rate in high-viscosity fluids is more significant due to flow instability.

[0108] Exception handling mechanisms include: If the maximum value of the normalized parameter is equal to the minimum value (no data fluctuation), the normalization step is skipped and the original value is directly multiplied by the weight; if there is no matching entry for the current fluid type in the weight configuration table, the default weight (50%:50%) is used and an alarm is triggered. Hardware requirements include: a microcontroller with floating-point support (such as the STM32F746, 216MHz, 320KB of memory) for real-time weighted sum calculations; and software requirements include an open-source linear algebra library (such as Arm CMSIS-DSP 5.7.0).

[0109] There are two ways to obtain viscosity characteristics: one is real-time detection through an online viscosity sensor (such as a rotational viscometer with a measurement range of 1 cP to 10,000 cP and an accuracy of ±2% FS); the other is to call up historical viscosity data from a preset parameter library based on the fluid type. The preset parameter library stores typical viscosity values ​​for common fluids, such as water-based fluids (1 cP to 10 cP), crude oil (800 cP to 1200 cP), and polymer solutions (5000 cP to 10,000 cP). When using the parameter library mode, the user selects the fluid name (such as "water" or "crude oil") through the human-machine interface, and the system automatically matches the viscosity value.

[0110] The statistical rule for extreme value normalization is to use the rolling extreme values ​​of the same period over the past seven days for both maximum and minimum values. For example, data from 8:00 AM each day is counted separately to avoid confusion between data from different operating periods. If insufficient data is available within seven days (e.g., fewer than 10 data sets), the statistical period is expanded to 30 days to ensure extreme value reliability. Normalization results are rounded to two decimal places. Values ​​outside the range of 0 to 1 are truncated to 0 or 1, and an exception log is recorded.

[0111] Dynamic adjustment of weight distribution is achieved through a lookup table method: the system has a built-in weight mapping table, where each row corresponds to a fluid type and its viscosity range, and each column corresponds to weight A (complexity attenuation rate) and weight B (low-frequency energy proportion). For example, when the fluid type is "water", the viscosity range is 1-10cP, weight A = 30%, weight B = 70%; when the fluid type is "crude oil", the viscosity range is 800-1200cP, weight A = 65%, weight B = 35%. The lookup table data is generated through factory calibration experiments. The calibration method is: in a simulated blockage scenario, the detection accuracy of different weight combinations is tested separately, and the combination with the highest accuracy is selected and written into the table.

[0112] During the real-time calculation of the linear weighted sum, if an input parameter is abnormal (e.g., a normalized value greater than 1 or less than 0), the normalized value from the previous period is interpolated instead. The interpolation method is: if the current data is abnormal, the moving average of the data from the previous three periods is used as the replacement value. For example, if the current normalized complexity decay rate is 1.2 (abnormal), and the data from the previous three periods are 0.6, 0.7, and 0.8, the replacement value is (0.6 + 0.7 + 0.8) / 3 = 0.7.

[0113] Boundary condition handling includes switching to parameter library mode and employing a nearest neighbor matching strategy when the fluid viscosity exceeds the sensor's range (e.g., >10,000 cP). For example, if a viscosity of 15,000 cP is detected, the system matches the weight of the fluid type with the highest viscosity in the parameter library (e.g., "asphalt," which has a viscosity of 10,000 cP). If no matching entry is found in the parameter library, manual intervention mode is triggered, suspending automatic detection until an engineer manually configures the system.

[0114] When the entanglement risk coefficient exceeds the third threshold, the cleaning device is controlled to perform directional pulse flushing on the wedge-shaped area and recalibrate the zero point of the pressure difference signal after flushing. The specific implementation is as follows:

[0115] When the entanglement risk coefficient exceeds the third threshold, the specific implementation method of controlling the cleaning device to perform directional pulse flushing on the wedge-shaped area is as follows: the intensity and duration of the flushing pulse are dynamically adjusted according to the mapping relationship between the fluid type and the entanglement risk coefficient. The fluid type is obtained through a viscosity sensor or a preset parameter library. The flushing pulse intensity ranges from 0.5MPa to 5MPa, and the duration ranges from 1 second to 10 seconds. For example, when the fluid is detected to be high-viscosity crude oil and the entanglement risk coefficient is 0.8 (the third threshold is 0.6), the pulse intensity is set to 4MPa and the duration is 8 seconds; if the fluid is a low-viscosity water-based solution and the risk coefficient is 0.7, the intensity is adjusted to 2MPa and the duration is 5 seconds. The parameter mapping relationship is calibrated through experiments: in a simulated blockage scenario, the pulse intensity is gradually increased until the entanglement is cleared, and the minimum effective intensity and corresponding time are recorded to form an optimal parameter table for different fluid types.

[0116] The basis for setting the third threshold is: through the distribution statistics of the entanglement risk coefficient of normal working conditions and entanglement and blockage working conditions in historical data, the critical value for distinguishing the two types of working conditions is selected. The specific method is: collect at least 100 groups of historical data (50 groups of normal and 50 groups of blocked), draw a histogram of the risk coefficient distribution, and take the 95th percentile of the normal working condition distribution as the third threshold. For example, the maximum value of the normal working condition risk coefficient is 0.55, and the minimum value of the blocked working condition is 0.62, then the third threshold is set to 0.6. For different fluid types, the threshold is dynamically adjusted: the threshold for high viscosity fluid (>1000cP) is increased by 0.05 to 0.1, and the threshold for low viscosity fluid (<100cP) is reduced by 0.05 to 0.1. The threshold data is stored in the parameter library and supports calling according to the real-time fluid type. If there is no matching item in the parameter library, the default threshold of 0.6 is used and an alarm is triggered.

[0117] After flushing is complete, a preset delay window is set to allow fluid flow to stabilize. This delay window is determined by pipe diameter and fluid viscosity: 5 to 10 seconds for pipes smaller than DN100 and with viscosities less than 100 cP, and 20 to 30 seconds for pipes larger than DN200 or with viscosities greater than 1000 cP. For example, a DN150 pipe carrying a medium-viscosity fluid (500 cP) would have a 15-second delay. During this delay, the system disables differential pressure signal acquisition to prevent interference from fluid disturbances.

[0118] When calculating the zero offset by acquiring a stable differential pressure signal, the zero offset is calculated as the absolute difference between the current differential pressure signal mean and the zero reference value stored in the historical database. For example, if the historical reference value is -0.1 kPa and the current signal mean is 0.3 kPa, the zero offset is 0.4 kPa. The preset tolerance range is set based on the sensor accuracy: for a sensor accuracy of ±0.05% FS (±10 kPa full scale), the tolerance range is set to ±0.1 kPa. If the offset exceeds the tolerance, the system automatically marks the current calibration as failed.

[0119] The flushing and calibration process is repeated until the offset reaches the target. The maximum number of repetitions is set to three, and the delay after each flush is incremented by 10 seconds to avoid over-cleaning. For example, the delay after the first flush is 15 seconds, the second 25 seconds, and the third 35 seconds. If the target is still not reached after three flushes, a manual intervention alarm is triggered and automatic cleaning is suspended. After calibration, the zero reference value is updated as follows: the average of the zero points of the last three successful calibrations is written to the historical database, overwriting the previous reference value. The historical database uses a ring buffer to store the last 30 days of reference data for easy access.

[0120] The hardware required to dynamically adjust flushing parameters includes a high-pressure pulse pump (such as a plunger pump with a pressure range of 0.1 MPa to 10 MPa) and a flow control valve (with a response time of less than 0.5 seconds). Software requirements include a real-time control system (such as FreeRTOS 10.4.3) and a calibration algorithm library (such as the sliding mean function in NumPy 1.21.5). During the flushing process, if the pulse pump pressure sensor detects an abnormal pressure (exceeding 10% of the set value), the flushing process is immediately terminated and fault diagnosis is triggered.

[0121] A mapping table of fluid types and flushing parameters is generated by simulating blockage scenarios with different fluids (water, crude oil, and polymer solutions) in the laboratory. The pulse intensity is gradually increased until the blockage is completely cleared, and the minimum effective intensity and corresponding time are recorded. For example, for crude oil with a viscosity of 800 cP at 25°C, the minimum effective intensity is 3.8 MPa for 7 seconds; a polymer solution (5000 cP) requires a strength of 5.2 MPa for 12 seconds. The mapping table is stored in JSON format and contains four columns: fluid type, viscosity range, pulse intensity, and duration. It supports online updates.

[0122] The delay time window is set based on fluid dynamics simulation results: by calculating the Reynolds number and flow recovery time of the fluid in the pipe, the quantitative relationship between delay time, viscosity, and pipe diameter is determined. For example, when the Reynolds number is less than 2000 (laminar flow), delay time = pipe diameter (m) × 10 + viscosity (cP) × 0.02; when the Reynolds number is greater than 4000 (turbulent flow), delay time = pipe diameter (m) × 5 + viscosity (cP) × 0.01. The simulation model was verified using ANSYS Fluent, with an error rate of less than 5%.

[0123] To calculate the zero offset, the mean of the differential pressure signal is obtained using a sliding window method. The window length is set to 10 seconds after the end of the delay time. The mean is calculated once per second, and outliers within ±3 times the standard deviation are removed. For example, if 1000 differential pressure points are collected in 10 seconds, 950 valid points are retained after removal, with a mean of 0.05 kPa. If the percentage of valid data falls below 80%, the flow is considered unstable and the delay time is extended by 10 seconds.

[0124] The historical database's circular buffer management strategy is to automatically delete the oldest day's data at midnight each day, retaining the most recent 30 days. When the baseline value is updated, the system verifies the new value's plausibility. If the new baseline value deviates from the previous value by more than three times the sensor's accuracy (e.g., ±0.15 kPa), a manual verification process is triggered to prevent abnormal data from contaminating the database.

[0125] Hardware fault diagnosis logic includes the following: If the pulse pump pressure sensor exceeds the limit for three consecutive times, the system switches to the backup pump and records the fault code. If the flow control valve responds in a timeout (failure to reach the set opening for more than 0.5 seconds), the backup valve is activated and the opening position is calibrated. Fault data is uploaded to the monitoring center via the Modbus protocol, supporting remote diagnosis.

[0126] The technical improvements of this embodiment are reflected in the collaborative design of signal feature extraction, multi-dimensional fusion and dynamic control. Traditional anti-clogging methods usually rely on a single signal dimension (such as a pressure difference threshold) or a fixed cleaning logic, making it difficult to distinguish between entanglement-type blockages and particulate impurity interference, and lack the ability to adapt to the dynamic characteristics of the fluid. This embodiment constructs a dual criterion for entanglement risk through the cross-dimensional coupling of frequency domain energy distribution (low-frequency energy proportion) and time domain complexity attenuation (sample entropy multi-scale analysis). The frequency domain features capture the energy concentration effect of flow field distortion, and the time domain features quantify the dynamic evolution rate of flow instability. The two are nonlinearly integrated through adaptive weight distribution of fluid viscosity, breaking through the limitations of traditional linear weighting or independent criteria. Furthermore, the cleaning control logic does not simply respond to threshold violations, but dynamically adjusts the pulse parameters according to the fluid type and risk level, and combines delay stabilization and closed-loop calibration mechanisms to solve the problem of false detection caused by flow field disturbances after a single flush. The synergistic effect of the above steps enables the system to distinguish the differential signal characteristics of fiber entanglement and particle scouring, while overcoming the inherent interference of signal lag in high-viscosity fluids.

[0127] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to actual conditions.

[0128] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0129] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0130] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0131] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0132] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0133] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An anti-clogging measurement and control system for a wedge-shaped flowmeter, characterized in that: include: A pressure difference detection module is used to detect the pressure difference signal between the inlet and outlet of the wedge flowmeter and generate a corresponding frequency spectrum distribution diagram; Feature extraction module, used to extract the low-frequency energy ratio of flow field distortion and the number of high-frequency noise mutation points from the spectrum distribution diagram; A trigger determination module is used to determine whether to trigger winding type congestion analysis based on whether the proportion of low-frequency energy exceeds a first threshold and the number of high-frequency noise mutation points is less than a second threshold; The entropy analysis module is used to perform multi-scale sample entropy analysis on the pressure difference signal when the winding type blockage analysis is triggered, fit the entropy attenuation slope of each scale, and generate the complexity attenuation rate; The risk fusion module is used to generate the entanglement risk coefficient based on the complexity decay rate and the proportion of low-frequency energy; The control execution module is used to control the cleaning device to perform directional pulse flushing on the wedge-shaped area when the entanglement risk coefficient exceeds a third threshold value, and to recalibrate the zero point of the pressure difference signal after flushing.

2. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: Detect the pressure difference signal between the inlet and outlet of the wedge flowmeter and generate the corresponding spectrum distribution diagram, including: The real-time differential pressure signal between the inlet and outlet of the wedge flowmeter is continuously collected through the pressure sensor; preprocessing the pressure difference signal to remove high-frequency noise and baseline drift, thereby generating a preprocessed pressure difference signal; Based on the preprocessed pressure difference signal, a frequency spectrum distribution diagram covering a preset frequency band is generated through fast Fourier transform.

3. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: Extract the low-frequency energy ratio of flow field distortion and the number of high-frequency noise mutation points from the spectrum distribution diagram, including: Integrate and calculate the normalized amplitude energy in a preset low-frequency band of the spectrum distribution diagram to obtain a low-frequency energy value; Integrate and calculate the normalized amplitude energy within a preset high frequency band of the spectrum distribution diagram to obtain a high frequency energy value; The ratio of the low-frequency energy value to the sum of the low-frequency energy value and the high-frequency energy value is taken as the low-frequency energy proportion; The number of amplitude energy differences between adjacent frequency points in the high-frequency band exceeding the preset mutation threshold is detected and counted as the number of high-frequency noise mutation points.

4. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: Whether to trigger winding type congestion analysis is determined based on whether the low-frequency energy ratio exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold, including: Compare the low-frequency energy ratio with a preset first threshold, and compare the number of high-frequency noise mutation points with a preset second threshold; When the proportion of low-frequency energy exceeds the first threshold and the number of high-frequency noise mutation points is less than the second threshold, the winding type congestion analysis is triggered, otherwise it is not triggered.

5. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 4, characterized in that: The first threshold is dynamically adjusted according to a preset multiple range of the mean low-frequency energy proportion under historical normal operating conditions corresponding to the fluid type; the second threshold is set according to the high quantile of the statistical distribution of the number of high-frequency noise mutation points under non-blocking conditions.

6. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: When triggering the winding type blockage analysis, multi-scale sample entropy analysis is performed on the pressure difference signal, and the attenuation slope of the entropy value at each scale is fitted to generate the complexity attenuation rate, including: Based on the dynamic fluctuation characteristics of the pressure difference signal, the coarse-graining time scale range is divided according to the fluid turbulence intensity; Generate the corresponding coarse-grained sequence according to the coarse-grained time scale range; Calculate the sample entropy value for each coarse-grained sequence and detect its non-stationary decay trend with increasing scale; In view of the non-stationary attenuation trend, piecewise linear regression is used to fit the attenuation slopes of the initial segment and the stable segment respectively; the absolute value of the difference between the attenuation slopes of the initial segment and the stable segment is taken as the complexity attenuation rate.

7. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 6, characterized in that: Low turbulence intensity corresponds to long-scale sequences, and high turbulence intensity corresponds to short-scale sequences.

8. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: Based on the complexity attenuation rate and the proportion of low-frequency energy, the entanglement risk coefficient is generated, including: Dynamically allocate the fusion weight of complexity attenuation rate and low-frequency energy proportion according to the viscosity characteristics corresponding to the fluid type; The complexity attenuation rate and the low-frequency energy ratio are normalized to eliminate the dimensional differences. The normalized complexity decay rate and the low-frequency energy ratio are linearly weighted and summed according to the assigned weights to generate the entanglement risk coefficient; When the sum of the fusion weights exceeds the preset range, they are redistributed according to the weight ratio until the sum reaches the set value.

9. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 8, characterized in that: The weight of the complexity decay rate corresponding to high viscosity fluid increases.

10. The anti-clogging measurement and control system for a wedge-shaped flowmeter according to claim 1, characterized in that: When the entanglement risk coefficient exceeds the third threshold, the cleaning device is controlled to perform directional pulse flushing on the wedge-shaped area, and the zero point of the pressure difference signal is recalibrated after flushing, including: When the entanglement risk coefficient exceeds a third threshold, the intensity and duration of the flushing pulse are dynamically adjusted according to a mapping relationship between the fluid type and the entanglement risk coefficient; After the flushing is completed, a preset time window is delayed to wait for the fluid flow to stabilize, and then the differential pressure signal in the stable state is collected to calculate the zero offset; If the zero offset exceeds the preset tolerance range, repeat the flushing and calibration process until the offset meets the standard; After calibration is completed, the zero point reference value in the historical database is updated for subsequent detection cycles.

Images