An environmental pollution online monitoring method and system

Abstract

This invention relates to the field of environmental pollution monitoring technology, specifically an online environmental pollution monitoring method and system. The method involves real-time acquisition of raw acoustic signals from the inner and outer walls of a bend in a high-risk monitoring area, followed by propagation path compensation to obtain a set of acoustic characteristic parameters for the inner and outer sides. The thickness and adhesion strength of the biofilm on both sides of the bend are then analyzed. The method calculates the detachment risk factors on both sides and a comprehensive detachment risk value. When the comprehensive detachment risk value exceeds a preset detachment risk threshold, a biofilm detachment risk warning is generated. This effectively solves the monitoring problem of the heterogeneous spatial distribution of biofilm at bends in pipeline networks, achieving accurate early warning of biofilm detachment risk in bend areas.

Description

Technical Field

[0001] This invention relates to the field of environmental pollution monitoring technology, specifically to an online environmental pollution monitoring method and system. Background Technology

[0002] In the field of environmental pollution monitoring, monitoring the biofilm status of pipeline systems is a crucial step in ensuring water quality safety and the healthy operation of pipelines. At bends in pipelines, the fluid dynamics change significantly. The curvature of the bend creates a low-velocity zone on the inside and a high-velocity zone on the outside. This velocity gradient causes spatial differences in centrifugal and shear forces, resulting in uneven biofilm thickness and heterogeneous adhesion strength on both sides of the bend. Abnormal biofilm shedding can carry heavy metals or organic pollutants into water bodies, posing a potential threat to the ecological environment and public health.

[0003] However, in current pipeline monitoring using acoustic sensors, the acoustic signals are distorted due to the curvature structure during propagation in bends, making it difficult to capture the differentiated characteristics on the inside and outside of the bend. This results in the inability to effectively obtain spatial distribution data of biofilm thickness and adhesion strength on the inside and outside of the bend, and consequently, the inability to accurately analyze the risk of biofilm detachment.

[0004] To address the aforementioned technical bottlenecks, this invention proposes an online environmental pollution monitoring method that effectively solves the problem of monitoring the spatial heterogeneity of biofilm distribution at pipe bends, enabling precise early warning of biofilm shedding risks in bend areas. Summary of the Invention

[0005] (1) Technical problems to be solved

[0006] The purpose of this invention is to provide an online environmental pollution monitoring method and system to solve the problem of inaccurate monitoring of the spatial heterogeneity of biofilm distribution caused by fluid dynamic asymmetry at pipe bends, which makes it impossible to achieve accurate early warning of the risk of biofilm shedding in bend areas.

[0007] (2) Technical solution

[0008] To achieve the above objectives, in one aspect, the present invention provides an online environmental pollution monitoring method, the method comprising:

[0009] S1. Identify the bend structure data in the pipeline network based on the pipeline geographic information topology data, obtain the curvature radius and pipe diameter of each bend based on the bend structure data, and mark the bends with curvature radius less than or equal to a preset multiple of pipe diameter as high-risk monitoring areas.

[0010] S2. Deploy an array of acoustic sensors in the high-risk monitoring area to collect the original acoustic signals of the inner and outer walls of the curve in real time; perform propagation path compensation on the original acoustic signals according to the radius of curvature and sound wave frequency of the high-risk monitoring area to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

[0011] S3. The inner acoustic feature parameter set and the outer acoustic feature parameter set are analyzed to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend.

[0012] S4. Calculate the centrifugal force enhancement coefficient based on the radius of curvature of the bend, and calculate the inner detachment risk factor based on the centrifugal force enhancement coefficient, the inner biofilm thickness value, and the inner adhesion strength value; obtain the shear force sensitivity coefficient of the bend and calculate the outer detachment risk factor based on the outer biofilm thickness value and the outer adhesion strength value.

[0013] S5. Calculate the comprehensive detachment risk value based on the inner and outer detachment risk factors; when the comprehensive detachment risk value exceeds the preset detachment risk threshold, generate a biofilm detachment risk warning message.

[0014] Furthermore, the method for obtaining the inner acoustic feature parameter set and the outer acoustic feature parameter set by compensating the propagation path of the original acoustic signal based on the radius of curvature and sound wave frequency of the high-risk monitoring area includes:

[0015] Obtain the bending angle of the curve in the curve structure data of the high-risk monitoring area, calculate the equivalent sound path difference between the inner and outer sound wave propagation paths based on the curvature radius, pipe diameter and bending angle of the curve, and analyze the amplitude compensation coefficient and time delay correction amount based on the equivalent sound path difference.

[0016] The original acoustic signal is decomposed into an inner reflected wave signal component and an outer reflected wave signal component according to the direction of arrival estimation method, and the corresponding amplitude compensation coefficient and time delay correction amount are applied to obtain the compensated inner reflected wave signal component and outer reflected wave signal component.

[0017] The compensated inner and outer reflected wave signal components are subjected to time-frequency domain feature extraction to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

[0018] Furthermore, the method for obtaining the amplitude compensation coefficient and time delay correction amount based on the equivalent sound path difference analysis includes:

[0019] Obtain the current pipe material type of the corresponding pipe bend in the high-risk monitoring area, obtain the propagation speed of sound waves in the pipe wall medium based on the current pipe material type, and convert the equivalent sound path difference into a time delay correction amount based on the propagation speed.

[0020] The attenuation base value, frequency index, and frequency-dependent attenuation coefficient corresponding to the current pipe type are obtained, and the amplitude compensation coefficient is calculated by combining the equivalent sound path difference and the sound wave frequency component. The attenuation base value, frequency index, and frequency-dependent attenuation coefficient are pre-determined through calibration experiments and stored as a mapping table for different pipe types.

[0021] Furthermore, the method for pre-determining the frequency-dependent attenuation coefficient through calibration experiments includes:

[0022] An array of reference acoustic sensors is deployed upstream of the straight pipe section in the high-risk monitoring area to collect reference acoustic signals without curve distortion in real time; the difference in attenuation slope between the original acoustic signal and the reference acoustic signal at the curve in the high-risk monitoring area is calculated in the same characteristic frequency band; a dynamic correction factor is generated based on the difference in attenuation slope, and the frequency-dependent attenuation coefficient is updated based on the dynamic correction factor.

[0023] Further, the method of generating a dynamic correction factor based on the attenuation slope difference and updating the frequency-dependent attenuation coefficient based on the dynamic correction factor includes:

[0024] The system acquires a set of current pipeline environmental parameters in real time, including water temperature, fluid turbidity, and flow velocity; and generates an environmental impact weight matrix based on the set of environmental parameters.

[0025] A dynamic correction factor is calculated based on the attenuation slope difference and the environmental influence weight matrix; the dynamic correction factor is multiplied by the preset learning rate coefficient to obtain the final correction step size; the final correction step size is superimposed on the current frequency-dependent attenuation coefficient to obtain the updated frequency-dependent attenuation coefficient.

[0026] Furthermore, the method for extracting time-frequency domain features from the compensated inner reflected wave signal component and the outer reflected wave signal component to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set includes:

[0027] The compensated inner and outer reflected wave signal components are subjected to short-time Fourier transform to obtain a time-spectrum matrix, and the power spectral density distribution curve is calculated based on the time-spectrum matrix within a preset characteristic frequency band.

[0028] The acoustic reflection coefficient of the biofilm-pipe wall interface is calculated based on the power spectral density distribution curve, and an acoustic impedance spectrum is generated based on the acoustic reflection coefficient. Within the preset characteristic frequency band of the time-frequency matrix, the spectral attenuation curves at each time point are extracted along the frequency axis. An envelope is extracted from the spectral attenuation curve at each time point, and the attenuation slope parameter is generated by fitting the slope of the envelope through linear regression. The center frequency of the resonance peak of the power spectral density distribution curve is detected within the preset characteristic frequency band, and the resonance frequency shift is calculated by comparing the center frequency of the resonance peak with the standard resonance frequency of the same pipe type without biofilm coverage.

[0029] The acoustic impedance spectrum, attenuation slope parameter, and resonance frequency shift constitute the inner acoustic characteristic parameter set and the outer acoustic characteristic parameter set.

[0030] Furthermore, the method for analyzing the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value, and outer biofilm thickness value and outer adhesion strength value of the curve includes:

[0031] The initial estimate of biofilm porosity is calculated based on the ratio of the real to the imaginary part of the acoustic impedance spectrum; the intermediate value of biofilm viscoelastic modulus is obtained by using the attenuation slope parameter through a pre-stored viscoelastic mapping table; the biofilm density correction factor is calculated based on the resonance frequency shift and the initial estimate of porosity; the biofilm structural stiffness matrix is ​​generated by coupling the intermediate value of biofilm viscoelastic modulus and the biofilm density correction factor; and the biofilm thickness and adhesion strength are obtained by solving the biofilm structural stiffness matrix.

[0032] Furthermore, the method for obtaining the biofilm thickness and adhesion strength values ​​by solving the biofilm structural stiffness matrix includes:

[0033] The stiffness matrix of the biofilm structure is decomposed into a symmetric positive definite matrix, and the eigenvector corresponding to its minimum eigenvalue is solved by the inverse power iteration method. The eigenvector is then mapped to the initial solution of biofilm thickness and the initial solution of adhesion strength through dimension conversion coefficients.

[0034] The dimensionless deviation is calculated based on the initial solution of biofilm thickness and the initial solution of adhesion strength. When the deviation exceeds the preset deviation threshold, the deviation is introduced as a penalty term into the objective function of eigenvalue residual for iteration until the deviation obtained by iteration does not exceed the preset deviation threshold and the eigenvalue residual is less than the preset residual threshold. Then, the final solution of biofilm thickness and adhesion strength is output and recorded as biofilm thickness value and adhesion strength value.

[0035] Furthermore, the method of incorporating the deviation as a penalty term into the eigenvalue residual objective function for iteration includes:

[0036] Obtain the dimensionless deviation of the current iteration and multiply it by a preset penalty weight factor to generate a penalty term; obtain the objective function of the eigenvalue residual of the current iteration based on the penalty term and the eigenvalue residual obtained in the previous iteration.

[0037] The objective function of the eigenvalue residuals is solved by the gradient descent algorithm to obtain a new eigenvector; the updated initial solutions for biofilm thickness and adhesion strength are extracted from the new eigenvector and iteratively calculated.

[0038] On the other hand, based on the same inventive concept, this invention also provides an online environmental pollution monitoring system, which includes: a high-risk monitoring area identification module, an acoustic signal acquisition and compensation module, an acoustic feature parameter analysis module, a shedding risk factor calculation module, and a shedding risk early warning generation module, with each module being connected in a sequential communication manner;

[0039] The high-risk monitoring area identification module is used to identify the bend structure data in the pipeline network based on the pipeline geographic information topology data, obtain the curvature radius and pipe diameter of each bend based on the bend structure data, and mark the bends with curvature radius less than or equal to a preset multiple of pipe diameter as high-risk monitoring areas.

[0040] The acoustic signal acquisition and compensation module is used to deploy an array of monitoring acoustic sensors in the high-risk monitoring area to acquire the original acoustic signals of the inner and outer walls of the curve in real time; and to perform propagation path compensation on the original acoustic signals according to the radius of curvature and sound wave frequency of the high-risk monitoring area to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

[0041] The acoustic feature parameter analysis module is used to analyze the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend.

[0042] The detachment risk factor calculation module is used to calculate the centrifugal force enhancement coefficient based on the radius of curvature of the bend, and to calculate the inner detachment risk factor based on the centrifugal force enhancement coefficient, the inner biofilm thickness value, and the inner adhesion strength value; and to obtain the shear force sensitivity coefficient of the bend and calculate the outer detachment risk factor by combining it with the outer biofilm thickness value and the outer adhesion strength value.

[0043] The detachment risk warning generation module is used to calculate a comprehensive detachment risk value based on the inner and outer detachment risk factors; when the comprehensive detachment risk value exceeds a preset detachment risk threshold, a biofilm detachment risk warning message is generated.

[0044] (3) Beneficial effects

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] 1. High-risk monitoring areas are accurately identified by using pipeline geographic information topology data, and acoustic sensor arrays are deployed to collect raw acoustic signals in real time. Then, the propagation path of the signals is compensated by combining the radius of curvature and sound wave frequency to obtain a precise set of acoustic characteristic parameters of the inner and outer sides. This enables high-precision and efficient monitoring of the biofilm status on the inner wall of the pipeline bends, ensuring that the monitoring targets cover the areas in the pipeline network with the highest risk of detachment.

[0047] 2. In obtaining the amplitude compensation coefficient and time delay correction, the influence of different pipe material types is considered. Relevant parameters are pre-determined through calibration experiments and stored as a mapping table. Based on the set of pipeline environmental parameters, a dynamic correction factor is generated in real time to update the frequency-dependent attenuation coefficient, enabling the monitoring method to adapt to different pipeline environments and pipe material conditions, thereby improving the accuracy and adaptability of monitoring.

[0048] 3. The biofilm thickness and adhesion strength values ​​are obtained by analyzing the acoustic characteristic parameter sets of the inner and outer sides. A detachment risk factor is calculated and an early warning is issued by comprehensively considering the centrifugal force enhancement coefficient and the shear force sensitivity coefficient. In the process of analyzing biofilm parameters, methods such as solving the biofilm structural stiffness matrix and introducing penalty terms for iteration are used to further improve the accuracy of biofilm parameter analysis, thereby enabling more accurate and timely generation of biofilm detachment risk warning information. Attached Figure Description

[0049] Figure 1 This is a flowchart of an online environmental pollution monitoring method according to Embodiment 1 of the present invention.

[0050] Figure 2 This is a schematic diagram of the module composition of an online environmental pollution monitoring system according to Embodiment 2 of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Before giving examples, it is necessary to describe the application scenarios of the present invention. The present invention is an online monitoring method and system for environmental pollution. It is applied to the problem that the biofilm in the bend of the pipeline network has uneven thickness and heterogeneous adhesion strength due to the significant change in the fluid dynamic characteristics. Abnormal shedding of the biofilm can lead to the entry of heavy metals or organic pollutants into the water body, which poses a potential threat to the ecological environment and public health.

[0053] Example 1: As Figure 1 As shown in the figure, this embodiment provides an online environmental pollution monitoring method, the method comprising:

[0054] S1. Identify the bend structure data in the pipeline network based on pipeline geographic information topology data. Obtain the radius of curvature and pipe diameter of each bend based on the bend structure data. Bends with a radius of curvature less than or equal to a preset multiple of the pipe diameter are marked as high-risk monitoring areas. Due to their special geometric structure, pipeline bends can cause significant changes in hydrodynamic characteristics, forming high-risk areas for biofilm detachment. All bend structures in the entire pipeline network are automatically identified by retrieving pipeline geographic information topology data. The bend structure data records the three-dimensional coordinates, pipe diameter specifications, and bending parameters of each bend in detail. For each identified bend, the radius of curvature and corresponding pipe diameter parameters are extracted. The smaller the radius of curvature, the more obvious the velocity gradient and centrifugal force effect at the bend, the greater the difference in biofilm distribution on the inner and outer walls, and the higher the risk of detachment. The preset multiple is set through experiments or simulations.

[0055] S2. Deploy an acoustic sensor array in the high-risk monitoring area to collect raw acoustic signals from the inner and outer walls of the bend in real time. Based on the radius of curvature and sound frequency of the high-risk monitoring area, perform propagation path compensation on the raw acoustic signals to obtain inner and outer acoustic characteristic parameter sets. The inner wall of the bend is divided into inner and outer walls according to the bend characteristics; the wall closer to the center of curvature is the inner wall, and the wall farther from the center of curvature is the outer wall. The acoustic sensor array is arranged in a ring around the pipe, capable of collecting acoustic signals from the inner and outer walls of the bend respectively. The advantage of acoustic monitoring technology lies in its non-invasiveness and real-time nature, allowing continuous acquisition of information on changes in the state of the biofilm on the pipe wall without affecting the normal operation of the pipeline. The sound frequency refers to the frequency of the sound waves actively emitted by the sensor array. Its selection needs to consider the pipe material, pipe diameter, fluid medium, and biofilm characteristics. Sound waves of different frequencies will be attenuated to varying degrees during propagation. The selection of the sound frequency is verified through acoustic simulation or experiments to ensure sufficient penetration and reflection sensitivity.

[0056] S3. The inner acoustic feature parameter set and the outer acoustic feature parameter set are analyzed to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend.

[0057] S4. Calculate the centrifugal force enhancement coefficient based on the radius of curvature of the bend. Then, calculate the inner biofilm thickness and adhesion strength based on this centrifugal force enhancement coefficient to obtain the inner detachment risk factor. Obtain the shear force sensitivity coefficient of the bend and calculate the outer detachment risk factor by combining it with the outer biofilm thickness and adhesion strength. The centrifugal force enhancement coefficient reflects the degree to which centrifugal force enhances the risk of biofilm detachment on the inner side of the bend. The detachment risk factor under centrifugal force is assessed by combining the inner biofilm thickness and adhesion strength parameters. Although the flow velocity is lower on the inner side of the bend, the centrifugal force is more significant, and the biofilm is at risk of detachment under centrifugal force. Considering the centrifugal force enhancement coefficient, biofilm thickness, and adhesion strength comprehensively allows for a more accurate assessment of this risk. For example, the larger the centrifugal force enhancement coefficient, and the thicker and lower the adhesion strength of the inner biofilm, the higher the inner detachment risk factor, indicating a greater likelihood of biofilm detachment under centrifugal force. For the outer side of a bend, due to the higher flow velocity and concentrated shear force, the shear force sensitivity coefficient is a parameter related to the characteristics of the bend itself, reflecting the sensitivity of the outer side of the bend to shear force. Bends with different radii of curvature, pipe diameters, and other conditions have different shear force sensitivity coefficients.

[0058] S5. Calculate the comprehensive detachment risk value based on the inner and outer detachment risk factors. When the comprehensive detachment risk value exceeds the preset detachment risk threshold, a biofilm detachment risk warning is generated. The comprehensive detachment risk value is the maximum value or weighted sum of the inner and outer detachment risk factors. The biofilm detachment risk warning is used to notify maintenance personnel to take corresponding preventive or treatment measures, thereby avoiding potential water pollution caused by large-scale biofilm detachment. The preset detachment risk threshold is a critical value determined by simulating biofilm detachment behavior under different operating conditions (including bend curvature radius, flow velocity, water quality, biofilm type, etc.) in the laboratory and actual pipeline system, collecting a large amount of experimental data, and using statistical analysis methods (such as ROC curves and cluster analysis). The preset detachment risk threshold is usually taken as the risk value when significant biofilm detachment begins at a confidence level ≥95%, and a certain safety margin can be set based on actual engineering experience.

[0059] The method for obtaining the inner acoustic feature parameter set and the outer acoustic feature parameter set by compensating the propagation path of the original acoustic signal based on the radius of curvature and sound wave frequency of the high-risk monitoring area includes:

[0060] The bending angle of the bend in the high-risk monitoring area's bend structure data is obtained. Based on the bend's radius of curvature, pipe diameter, and bending angle, the equivalent path difference between the inner and outer sound wave propagation paths is calculated. The amplitude compensation coefficient and time delay correction are then obtained based on this equivalent path difference. In the bend structure, the actual propagation path of the sound wave from the emission point to the receiving point is no longer a straight line, but rather propagates along the curved surface of the pipe wall. This path change causes signal amplitude attenuation and time delay. The bending angle determines the actual difference in the propagation path length between the inner and outer sides of the bend. Combined with the known radius of curvature and pipe diameter, the equivalent path difference between the inner and outer sound wave propagation paths can be accurately calculated.

[0061] The original acoustic signal is decomposed into inner and outer reflected wave signal components using a direction-of-arrival (DOA) estimation method. Corresponding amplitude compensation coefficients and time delay corrections are then applied to obtain the compensated inner and outer reflected wave signal components. The spatial decomposition of the original acoustic signal employs advanced DOA estimation techniques. Due to the circular arrangement of the acoustic sensor array, the reflected wave components from the inside and outside of the curve are accurately distinguished by analyzing the phase difference and time difference of the signals received by each sensor. This allows for the separation of signal components with clear spatial directivity in complex acoustic environments. The compensation process requires applying corresponding amplitude amplification and time delay correction to the inner and outer signal components respectively. Amplitude compensation primarily addresses energy loss caused by the propagation path, ensuring that the inner and outer signals have comparable energy references. Time delay correction ensures synchronization of the inner and outer signals on the time axis, eliminating the time offset effect caused by geometric path differences.

[0062] The compensated inner and outer reflected wave signal components are subjected to time-frequency domain feature extraction to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

[0063] The method for obtaining the amplitude compensation coefficient and time delay correction amount based on the equivalent sound path difference analysis includes:

[0064] The current pipe material type of the corresponding bend in the pipeline network of the high-risk monitoring area is obtained. Based on this pipe material type, the propagation speed of sound waves in the pipe wall medium is calculated. The equivalent sound path difference is then converted into a time delay correction based on this propagation speed. The time delay correction is linearly and positively correlated with the equivalent sound path difference. Common pipe material types include ductile iron, steel pipes, and plastic pipes, each with its specific sound wave propagation speed, attenuation characteristics, and frequency response features. Accurate determination of the sound wave propagation speed is crucial for calculating the time delay correction, as even small speed errors can accumulate into significant time deviations over long propagation distances. Based on the determined sound wave propagation speed in the pipe wall medium, the additional time delay due to the extended actual propagation path, i.e., the time delay correction, can be calculated using the equivalent sound path difference. A larger equivalent sound path difference indicates a longer actual sound wave propagation path, resulting in a larger time delay correction; the two are linearly and positively correlated.

[0065] The attenuation base value, frequency index, and frequency-dependent attenuation coefficient corresponding to the current pipe type are obtained, and the amplitude compensation coefficient is calculated by combining the equivalent sound path difference and sound wave frequency components. The attenuation base value, frequency index, and frequency-dependent attenuation coefficient are pre-determined through calibration experiments and stored as a mapping table for different pipe types. Amplitude compensation coefficient The calculation formula is: ;in, For equivalent sound path difference, For sound wave frequency components, This is the attenuation base value corresponding to the current pipe type. This is the frequency index corresponding to the current pipe type. This represents the frequency-dependent attenuation coefficient corresponding to the current pipe type. The acoustic frequency components refer to the frequency components obtained after spectral analysis of the acquired raw acoustic signal. Since different frequencies of sound waves have different attenuation characteristics when propagating in the pipe, the attenuation of each frequency component needs to be considered separately. The amplitude compensation coefficient is calculated using a more complex exponential attenuation model, which includes three key parameters: the attenuation base value reflects the basic attenuation characteristics of the pipe type, the frequency-dependent attenuation coefficient reflects the selective attenuation of different frequency sound waves by the pipe type, and the frequency exponent describes the nonlinear characteristics of this frequency dependence. The establishment of the mapping table is a systematic calibration process. Through numerous calibration experiments in the laboratory and in actual field conditions, the acoustic parameters of different pipe materials under various conditions are measured and organized into a database stored in the system, ensuring that the corresponding material parameters can be quickly and accurately retrieved during monitoring.

[0066] The method for pre-determining the frequency-dependent attenuation coefficient through calibration experiments includes:

[0067] A reference acoustic sensor array is deployed upstream of a straight pipe section in the high-risk monitoring area to acquire reference acoustic signals without bend distortion in real time. The difference in attenuation slope between the original acoustic signal and the reference acoustic signal at the bend in the high-risk monitoring area is calculated in the same characteristic frequency band. A dynamic correction factor is generated based on this difference in attenuation slope, and the frequency-dependent attenuation coefficient is updated based on this dynamic correction factor. The deployment strategy of the reference acoustic sensor array is crucial; it is installed on a straight pipe section upstream of the monitored bend, ensuring the standardization and controllability of the measurement environment. Sound wave propagation in the straight pipe section is unaffected by the geometric distortion of the bend, providing ideal reference measurement data. The reference acoustic sensor array and the monitoring acoustic sensor array use the same technical configuration, ensuring the comparability of measurement results. The analysis of the reference acoustic signal provides a standard reference for attenuation characteristics. The sound wave propagation path in the straight pipe section is a standard straight line, and the attenuation law follows classical acoustic theory. The reference attenuation slope spectrum represents the acoustic attenuation characteristics of the pipe type under ideal conditions. Comparative analysis of measured acoustic signals and reference acoustic signals at bends reveals the actual impact of bend geometry on acoustic propagation. The calculation of the attenuation slope difference not only considers the difference in average attenuation level but also analyzes the changes in attenuation characteristics at different frequency bands.

[0068] The method of generating a dynamic correction factor based on the attenuation slope difference and updating the frequency-dependent attenuation coefficient based on the dynamic correction factor includes:

[0069] The system acquires real-time sets of current pipeline environmental parameters, including water temperature, fluid turbidity, and flow velocity. An environmental impact weight matrix is ​​generated based on this parameter set. The water temperature weight factor is positively correlated with the absolute value of temperature change, the turbidity weight factor is linearly positively correlated with the turbidity value, and the flow velocity weight factor is exponentially correlated with the percentage deviation of the flow velocity from the design value. Water temperature monitoring reflects the impact of temperature changes on acoustic propagation characteristics. Temperature changes directly affect the sound velocity and attenuation characteristics of the pipe wall material, and also affect the acoustic properties of the fluid within the pipeline. The positive correlation between the water temperature weight factor and the magnitude of temperature change ensures the system's sensitive response to temperature influences. When the water temperature deviates significantly from the design value, the weight of the temperature factor in parameter correction is increased accordingly. Fluid turbidity monitoring reflects the impact of water quality on the acoustic environment. Increased turbidity means a higher concentration of suspended particles in the water. These particles scatter and absorb sound waves. The linear positive correlation of the turbidity weight factor reflects the directness and cumulative nature of this impact. In high-turbidity environments, the interpretation of acoustic signals requires greater correction. Flow velocity monitoring is a key indicator for assessing changes in shear environment and dynamic conditions. Deviations in flow velocity from design values ​​not only affect the growth and shedding characteristics of biofilms but also alter the propagation conditions of sound waves in the flowing medium. The flow velocity weighting factor is designed using an exponential relationship, reflecting the nonlinear characteristics of the impact of flow velocity changes on the acoustic environment. Significant changes in flow velocity cause a rapid increase in the weighting factor, ensuring the system's sensitive response to changes in hydraulic conditions. The construction of the environmental impact weighting matrix employs a multi-parameter fusion strategy, considering the interactions and coupling effects between parameters through matrix operations.

[0070] A dynamic correction factor is calculated based on the attenuation slope difference and the environmental impact weight matrix. The dynamic correction factor is multiplied by a preset learning rate coefficient to obtain the final correction step size. This final correction step size is then added to the current frequency-dependent attenuation coefficient to obtain the updated frequency-dependent attenuation coefficient. Matrix operations, such as weighted summation or matrix multiplication, are performed on the attenuation slope difference and the environmental impact weight matrix to obtain the dynamic correction factor. The dynamic correction factor integrates the effects of curved road structure and environmental disturbances. The learning rate coefficient controls the magnitude of the correction step size. Scaling the correction magnitude using the learning rate coefficient ensures the stability and convergence of the correction process. A smaller learning rate coefficient avoids drastic parameter oscillations, while a moderate adjustment ensures timely response to environmental changes. The preset learning rate coefficient controls the update step size of the frequency-dependent attenuation coefficient, typically ranging from 0.01 to 0.3. Real-time monitoring is conducted by selecting a portion of the pipeline in an actual pipeline monitoring site as a pilot area and setting different learning rate coefficients. The accuracy and stability of the monitoring results under different learning rate coefficients are compared, and the optimal learning rate coefficient is selected as the preset learning rate coefficient based on the actual monitoring data.

[0071] The method for extracting time-frequency domain features from the compensated inner reflected wave signal components and outer reflected wave signal components to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set includes:

[0072] The compensated inner and outer reflected wave signal components are subjected to short-time Fourier transform to obtain a time-spectrum matrix. Based on this time-spectrum matrix, a power spectral density distribution curve is calculated within a preset characteristic frequency band. Time-frequency domain feature extraction is a crucial step in converting the compensated acoustic signal into feature data usable for biomembrane parameter analysis. Short-time Fourier transform technology can simultaneously maintain signal resolution in both time and frequency, which is particularly important for analyzing non-stationary biomembrane acoustic signals. Changes in biomembrane state often manifest as temporal evolution of signal characteristics; neither simple frequency domain analysis nor time domain analysis can fully capture this dynamic characteristic. The time-spectrum matrix provides a rich information foundation for subsequent feature parameter extraction. The preset characteristic frequency band automatically determines the most effective analysis frequency band based on the fundamental frequency and harmonic components of the emitted sound wave, ensuring that the feature extraction process focuses on the frequency range with the richest information content. The calculation of the power spectral density distribution curve provides information on the distribution of acoustic energy in the frequency domain, reflecting the degree of influence of the biomembrane on sound waves of different frequencies. Accurate calculation of the power spectral density requires consideration of the signal's statistical characteristics and noise background, and the reliability of the results is ensured through a spectral estimation algorithm.

[0073] The acoustic reflection coefficient of the biofilm-pipe wall interface is calculated based on the power spectral density distribution curve, and an acoustic impedance spectrum is generated based on the acoustic reflection coefficient. Within a preset characteristic frequency band of the time-frequency matrix, spectral attenuation curves are extracted along the frequency axis at each time point. An envelope is extracted from the spectral attenuation curve at each time point, and the slope of the envelope is fitted using linear regression to generate an attenuation slope parameter. The center frequency of the resonance peak of the power spectral density distribution curve is detected within the preset characteristic frequency band, and the resonance frequency shift is calculated by comparing the center frequency of the resonance peak with the standard resonance frequency of the same pipe type without biofilm coverage. Differences in the acoustic properties of the biofilm-pipe wall interface cause partial reflection of sound waves, and the magnitude of the acoustic reflection coefficient is directly related to the physical properties of the biofilm. By comparing the power spectrum differences with and without biofilm, the reflection coefficient can be accurately calculated, thereby generating an acoustic impedance spectrum describing the acoustic properties of the interface. The attenuation slope parameter is extracted using a multi-time-point statistical analysis method. Spectral attenuation curves are continuously extracted throughout the monitoring period, and stable and reliable attenuation slope parameters are obtained through mathematical methods such as envelope fitting and linear regression. This time-averaging method effectively suppresses the influence of transient noise and improves the accuracy of parameter estimation. The detection of resonant frequency shift is based on the mechanism by which biofilms affect the resonant characteristics of the pipe wall. Biofilm adhesion alters the effective mass and stiffness characteristics of the pipe wall, causing a shift in the resonant frequency. By accurately measuring the resonant frequency shift and comparing it with a standard resonant frequency, the presence of the biofilm and its degree of influence can be confirmed. The resonant frequency shift detection technology has high sensitivity and can detect the presence of trace amounts of biofilm.

[0074] The acoustic impedance spectrum, attenuation slope parameter, and resonance frequency shift constitute the inner acoustic characteristic parameter set and the outer acoustic characteristic parameter set; the preset characteristic frequency band range is dynamically set according to the fundamental frequency and harmonic components of the emitted sound wave.

[0075] The method for analyzing the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value, and outer biofilm thickness value and outer adhesion strength value of the curve includes:

[0076] An initial estimate of biofilm porosity is calculated based on the ratio of the real to imaginary parts of the acoustic impedance spectrum. The attenuation slope parameter is used to obtain the intermediate value of the biofilm viscoelastic modulus through a pre-stored viscoelastic mapping table. A biofilm density correction factor is calculated based on the resonance frequency shift and the initial porosity estimate. The biofilm structural stiffness matrix is ​​generated by coupling the intermediate value of the biofilm viscoelastic modulus with the biofilm density correction factor. The pre-stored viscoelastic mapping table is established through calibration experiments on biofilm samples with different rheological properties. The biofilm thickness and adhesion strength values ​​are obtained by solving the biofilm structural stiffness matrix. The analysis of biofilm physical parameters is used to convert acoustic characteristic data into biofilm thickness and adhesion strength values ​​with clear physical meaning. The initial estimate of biofilm porosity utilizes the ratio of the real to imaginary parts in the acoustic impedance spectrum, reflecting the microstructural characteristics inside the biofilm, especially the relative proportions of pores and the solid matrix. Biofilm porosity is an important parameter affecting the acoustic properties of biofilms, influencing not only the propagation speed of sound waves but also the scattering and absorption characteristics of sound energy. The porosity estimates obtained through acoustic measurements provide crucial constraints for subsequent complex analyses. The establishment of a pre-stored viscoelastic mapping table is a systematic experimental calibration process. By culturing biofilm samples with different rheological properties, the correspondence between their acoustic attenuation characteristics and mechanical properties is determined. These experiments cover the full spectrum, from initially attached, thin biofilms to mature, thick biofilms, and various biofilm types formed under different microbial compositions and environmental conditions. The accuracy of the mapping table directly affects the reliability of parameter analysis. The calculation of the biofilm density correction factor embodies the idea of ​​multi-parameter coupled analysis. Both the resonant frequency shift and the initial porosity estimate contain density-related information, but reflect density characteristics from different physical perspectives. Coupled analysis of these two independent measurements yields a more accurate density correction factor, eliminating systematic errors that may arise from single-parameter analysis. The generation of the biofilm structural stiffness matrix is ​​a key step in converting acoustic measurement results into a mechanical description. The biofilm structural stiffness matrix integrates the viscoelastic characteristics, density distribution, and microstructure information of the biofilm, mathematically describing the response characteristics of the biofilm under various external forces.

[0077] The method for obtaining the biofilm thickness and adhesion strength values ​​by solving the biofilm structural stiffness matrix includes:

[0078] The stiffness matrix of the biofilm structure is decomposed into a symmetric positive definite matrix, and the eigenvectors corresponding to its minimum eigenvalues ​​are solved using the inverse power iteration method. These eigenvectors are then mapped to initial solutions for biofilm thickness and adhesion strength using dimension conversion coefficients. Advanced numerical analysis techniques are employed in the numerical solution process of the biofilm structure stiffness matrix to ensure the stability and convergence of the solutions. The symmetric positive definite decomposition of the biofilm structure stiffness matrix is ​​a crucial step in ensuring the stability of the numerical calculation. The symmetric positive definite mathematical property of the biofilm structure stiffness matrix reflects the fundamental physical characteristics of biofilm materials. Mathematical methods such as Cholesky decomposition transform the original matrix into a form more suitable for numerical calculation, avoiding potential numerical instability issues from direct solutions. The application of the inverse power iteration method reflects the need for accurate solutions to the minimum eigenvalues ​​and their corresponding eigenvectors. The minimum eigenvalues ​​correspond to the weakest response mode of the biofilm structure, which is closely related to the biofilm shedding mechanism. Iterative calculations can approximate the true eigenvalues ​​and eigenvectors with arbitrary precision. The application of dimension conversion coefficients solves the problem of the correspondence between mathematical solutions and physical parameters. The components of the eigenvector have abstract mathematical meanings and need to be mapped to physically meaningful thickness and adhesion strength values ​​using appropriate dimensional conversion coefficients. The determination of these conversion coefficients is based on extensive experimental calibration and theoretical analysis, ensuring the accuracy of the conversion process.

[0079] The dimensionless deviation is calculated based on the initial solutions for biofilm thickness and adhesion strength. When the deviation exceeds a preset deviation threshold, it is introduced as a penalty term into the eigenvalue residual objective function for iteration until the iterated deviation does not exceed the preset deviation threshold and the eigenvalue residual is less than the preset residual threshold. The final solutions for biofilm thickness and adhesion strength are then output and recorded as biofilm thickness and adhesion strength values. Since biofilm parameter analysis involves complex nonlinear relationships, the initial solution often requires iterative optimization to improve accuracy. The deviation calculation provides a quality assessment standard for the solution. When the deviation exceeds the preset deviation threshold, the iterative optimization program is initiated. The preset deviation threshold is used to judge the rationality of the initial solution for biofilm parameters and is typically taken as the 95th percentile of historical data or calibration deviations in experiments. The preset deviation threshold reflects the allowable error range for biofilm parameter analysis under normal monitoring conditions and can be determined through numerous calibration experiments, for example, 0.05–0.15. The preset residual threshold is used to control the accuracy of the iterative solution of eigenvalues. It is usually set according to numerical stability and computational resources, and the value is 1e-2 to 1e-4. It is necessary to ensure the accuracy of biofilm parameter analysis while avoiding excessive iteration that would lead to a decrease in computational efficiency.

[0080] The method of incorporating the deviation as a penalty term into the eigenvalue residual objective function for iteration includes:

[0081] The dimensionless deviation of the current iteration is obtained and multiplied by a preset penalty weight factor to generate a penalty term. The eigenvalue residual objective function of the current iteration is obtained based on the penalty term and the eigenvalue residual obtained in the previous iteration. The calculation and weight setting of the penalty term need to find a balance between mathematical convergence and physical constraints. The preset penalty weight factor is used to introduce deviation constraints into the eigenvalue residual objective function, with a value ranging from 0.1 to 1.0. The specific value can be determined through cross-validation or the L-curve method to achieve a balance between goodness of fit and constraint satisfaction. Too small a penalty weight factor may prevent the physical constraints from functioning effectively, while too large a penalty weight factor may cause difficulties in numerical calculation. The construction of the eigenvalue residual objective function comprehensively considers the dual requirements of mathematical solution accuracy and physical parameter constraints. Minimizing the eigenvalue residual objective function can simultaneously satisfy the solution requirements of the mathematical equations and the reasonableness constraints of the physical parameters.

[0082] The objective function of the eigenvalue residuals is solved using the gradient descent algorithm to obtain a new eigenvector. Updated initial solutions for biofilm thickness and adhesion strength are extracted from these new eigenvectors for iterative calculation. The application of the gradient descent algorithm provides an effective way to optimize the objective function. Gradient calculation reflects the sensitivity of the eigenvalue residual objective function to various variables, guiding the search direction of the iterative process. It not only checks whether the deviation meets the accuracy requirements but also monitors the changing trend of the eigenvalue residuals. When multiple convergence indicators simultaneously meet the requirements, the iterative process is considered to have converged, and the final optimization result is output.

[0083] Example 2: Based on the same inventive concept, such as Figure 2 As shown, this embodiment also provides an online environmental pollution monitoring system, which includes: a high-risk monitoring area identification module, an acoustic signal acquisition and compensation module, an acoustic feature parameter analysis module, a shedding risk factor calculation module, and a shedding risk early warning generation module, with each module connected in sequence via communication.

[0084] The high-risk monitoring area identification module is used to identify the bend structure data in the pipeline network based on the pipeline geographic information topology data, obtain the curvature radius and pipe diameter of each bend based on the bend structure data, and mark the bends with curvature radius less than or equal to a preset multiple of pipe diameter as high-risk monitoring areas.

[0085] The acoustic signal acquisition and compensation module is used to deploy an array of monitoring acoustic sensors in the high-risk monitoring area to acquire the original acoustic signals of the inner and outer walls of the curve in real time; and to perform propagation path compensation on the original acoustic signals according to the radius of curvature and sound wave frequency of the high-risk monitoring area to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

[0086] The acoustic feature parameter analysis module is used to analyze the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend.

[0087] The detachment risk factor calculation module is used to calculate the centrifugal force enhancement coefficient based on the radius of curvature of the bend, and to calculate the inner detachment risk factor based on the centrifugal force enhancement coefficient, the inner biofilm thickness value, and the inner adhesion strength value; and to obtain the shear force sensitivity coefficient of the bend and calculate the outer detachment risk factor by combining it with the outer biofilm thickness value and the outer adhesion strength value.

[0088] The detachment risk warning generation module is used to calculate a comprehensive detachment risk value based on the inner and outer detachment risk factors; when the comprehensive detachment risk value exceeds a preset detachment risk threshold, a biofilm detachment risk warning message is generated.

[0089] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0090] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for online monitoring of environmental pollution, characterized in that, The method includes: Based on the topological data of pipeline geographic information, identify the bend structure data in the pipeline network, obtain the curvature radius and pipe diameter of each bend based on the bend structure data, and mark the bends with curvature radius less than or equal to a preset multiple of pipe diameter as high-risk monitoring areas. An array of acoustic sensors is deployed in the high-risk monitoring area to collect the original acoustic signals of the inner and outer walls of the curve in real time; the propagation path compensation of the original acoustic signals is performed according to the radius of curvature and sound wave frequency of the high-risk monitoring area to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set; The inner acoustic feature parameter set and the outer acoustic feature parameter set are analyzed to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend; The centrifugal force enhancement coefficient is calculated based on the radius of curvature of the bend. The inner detachment risk factor is calculated based on the centrifugal force enhancement coefficient, the inner biofilm thickness value, and the inner adhesion strength value. The shear force sensitivity coefficient of the bend is obtained and combined with the outer biofilm thickness value and the outer adhesion strength value to calculate the outer detachment risk factor. A comprehensive detachment risk value is calculated based on the inner and outer detachment risk factors; when the comprehensive detachment risk value exceeds the preset detachment risk threshold, a biofilm detachment risk warning is generated.

2. The method for online monitoring of environmental pollution according to claim 1, characterized in that, The method for obtaining the inner acoustic feature parameter set and the outer acoustic feature parameter set by compensating the propagation path of the original acoustic signal based on the radius of curvature and sound wave frequency of the high-risk monitoring area includes: Obtain the bending angle of the bend in the bend structure data of the high-risk monitoring area, calculate the equivalent sound path difference between the inner and outer sound wave propagation paths based on the bend radius of curvature, pipe diameter and bending angle, and analyze the amplitude compensation coefficient and time delay correction amount based on the equivalent sound path difference. The original acoustic signal is decomposed into an inner reflected wave signal component and an outer reflected wave signal component according to the direction of arrival estimation method, and the corresponding amplitude compensation coefficient and time delay correction amount are applied to obtain the compensated inner reflected wave signal component and outer reflected wave signal component. The compensated inner and outer reflected wave signal components are subjected to time-frequency domain feature extraction to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set.

3. The method for online monitoring of environmental pollution according to claim 2, characterized in that, The method for obtaining the amplitude compensation coefficient and time delay correction amount based on the equivalent sound path difference analysis includes: Obtain the current pipe material type of the corresponding pipe bend in the high-risk monitoring area, obtain the propagation speed of sound waves in the pipe wall medium based on the current pipe material type, and convert the equivalent sound path difference into a time delay correction amount based on the propagation speed. The attenuation base value, frequency index, and frequency-dependent attenuation coefficient corresponding to the current pipe type are obtained, and the amplitude compensation coefficient is calculated by combining the equivalent sound path difference and the sound wave frequency component. The attenuation base value, frequency index, and frequency-dependent attenuation coefficient are pre-determined through calibration experiments and stored as a mapping table for different pipe types.

4. The method for online monitoring of environmental pollution according to claim 3, characterized in that, The method for pre-determining the frequency-dependent attenuation coefficient through calibration experiments includes: An array of reference acoustic sensors is deployed upstream of the straight pipe section in the high-risk monitoring area to collect reference acoustic signals without curve distortion in real time; the difference in attenuation slope between the original acoustic signal and the reference acoustic signal at the curve in the high-risk monitoring area is calculated in the same characteristic frequency band; a dynamic correction factor is generated based on the difference in attenuation slope, and the frequency-dependent attenuation coefficient is updated based on the dynamic correction factor.

5. The method for online monitoring of environmental pollution according to claim 4, characterized in that, The method of generating a dynamic correction factor based on the attenuation slope difference and updating the frequency-dependent attenuation coefficient based on the dynamic correction factor includes: The system acquires a set of current pipeline environmental parameters in real time, including water temperature, fluid turbidity, and flow velocity; and generates an environmental impact weight matrix based on the set of environmental parameters. A dynamic correction factor is calculated based on the attenuation slope difference and the environmental influence weight matrix; the dynamic correction factor is multiplied by the preset learning rate coefficient to obtain the final correction step size; the final correction step size is superimposed on the current frequency-dependent attenuation coefficient to obtain the updated frequency-dependent attenuation coefficient.

6. The method for online monitoring of environmental pollution according to claim 2, characterized in that, The method for extracting time-frequency domain features from the compensated inner reflected wave signal components and outer reflected wave signal components to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set includes: The compensated inner and outer reflected wave signal components are subjected to short-time Fourier transform to obtain a time-spectrum matrix, and the power spectral density distribution curve is calculated based on the time-spectrum matrix within a preset characteristic frequency band. The acoustic reflection coefficient of the biofilm-pipe wall interface is calculated based on the power spectral density distribution curve, and an acoustic impedance spectrum is generated based on the acoustic reflection coefficient. Within the preset characteristic frequency band of the time-frequency matrix, the spectral attenuation curves at each time point are extracted along the frequency axis. An envelope is extracted from the spectral attenuation curve at each time point, and the attenuation slope parameter is generated by fitting the slope of the envelope through linear regression. The center frequency of the resonance peak of the power spectral density distribution curve is detected within the preset characteristic frequency band, and the resonance frequency shift is calculated by comparing the center frequency of the resonance peak with the standard resonance frequency of the same pipe type without biofilm coverage. The acoustic impedance spectrum, attenuation slope parameter, and resonance frequency shift constitute the inner acoustic characteristic parameter set and the outer acoustic characteristic parameter set.

7. The method for online monitoring of environmental pollution according to claim 6, characterized in that, The method for analyzing the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value, and outer biofilm thickness value and outer adhesion strength value of the curve includes: The initial estimate of biofilm porosity is calculated based on the ratio of the real to the imaginary part of the acoustic impedance spectrum; the intermediate value of biofilm viscoelastic modulus is obtained by using the attenuation slope parameter through a pre-stored viscoelastic mapping table; the biofilm density correction factor is calculated based on the resonance frequency shift and the initial estimate of porosity; the biofilm structural stiffness matrix is ​​generated by coupling the intermediate value of biofilm viscoelastic modulus and the biofilm density correction factor; and the biofilm thickness and adhesion strength are obtained by solving the biofilm structural stiffness matrix.

8. The method for online monitoring of environmental pollution according to claim 7, characterized in that, The method for obtaining the biofilm thickness and adhesion strength values ​​by solving the biofilm structural stiffness matrix includes: The stiffness matrix of the biofilm structure is decomposed into a symmetric positive definite matrix, and the eigenvector corresponding to its minimum eigenvalue is solved by the inverse power iteration method. The eigenvector is then mapped to the initial solutions for biofilm thickness and adhesion strength through dimension conversion coefficients. The dimensionless deviation is calculated based on the initial solution of biofilm thickness and the initial solution of adhesion strength. When the deviation exceeds the preset deviation threshold, the deviation is introduced as a penalty term into the objective function of eigenvalue residual for iteration until the deviation obtained by iteration does not exceed the preset deviation threshold and the eigenvalue residual is less than the preset residual threshold. Then, the final solution of biofilm thickness and adhesion strength is output and recorded as biofilm thickness value and adhesion strength value.

9. The method for online monitoring of environmental pollution according to claim 8, characterized in that, The method of incorporating the deviation as a penalty term into the eigenvalue residual objective function for iteration includes: Obtain the dimensionless deviation of the current iteration and multiply it by a preset penalty weight factor to generate a penalty term; obtain the objective function of the eigenvalue residual of the current iteration based on the penalty term and the eigenvalue residual obtained in the previous iteration; The objective function of the eigenvalue residuals is solved by the gradient descent algorithm to obtain a new eigenvector; the updated initial solutions for biofilm thickness and adhesion strength are extracted from the new eigenvector and iteratively calculated.

10. An online environmental pollution monitoring system, characterized in that, The system includes: a high-risk monitoring area identification module, an acoustic signal acquisition and compensation module, an acoustic feature parameter analysis module, a shedding risk factor calculation module, and a shedding risk early warning generation module, with each module connected in sequence via communication. The high-risk monitoring area identification module is used to identify the bend structure data in the pipeline network based on the pipeline geographic information topology data, obtain the curvature radius and pipe diameter of each bend based on the bend structure data, and mark the bends with curvature radius less than or equal to a preset multiple of pipe diameter as high-risk monitoring areas. An acoustic signal acquisition and compensation module is used to deploy an array of acoustic sensors in the high-risk monitoring area to acquire the original acoustic signals of the inner and outer walls of the curve in real time; and to perform propagation path compensation on the original acoustic signals according to the radius of curvature and sound wave frequency of the high-risk monitoring area to obtain the inner acoustic feature parameter set and the outer acoustic feature parameter set. The acoustic feature parameter analysis module is used to analyze the inner acoustic feature parameter set and the outer acoustic feature parameter set to obtain the inner biofilm thickness value, inner adhesion strength value and outer biofilm thickness value and outer adhesion strength value of the bend. The detachment risk factor calculation module is used to calculate the centrifugal force enhancement coefficient based on the radius of curvature of the bend, and to calculate the inner detachment risk factor based on the centrifugal force enhancement coefficient, the inner biofilm thickness value, and the inner adhesion strength value; and to obtain the shear force sensitivity coefficient of the bend and combine it with the outer biofilm thickness value and the outer adhesion strength value to calculate the outer detachment risk factor. The detachment risk warning generation module is used to calculate a comprehensive detachment risk value based on the inner and outer detachment risk factors; when the comprehensive detachment risk value exceeds a preset detachment risk threshold, a biofilm detachment risk warning message is generated.

Images