In-situ nondestructive membrane pollution detection system and method based on ultrasonic transmission signal

Abstract

The invention belongs to the technical field of sewage treatment and membrane separation, and provides an in-situ nondestructive membrane pollution detection system and method based on an ultrasonic transmission signal, and the system comprises an ultrasonic excitation module which is used for generating a sine wave excitation signal which penetrates through a membrane body to excite the ultrasonic transmission signal for pollution detection; the signal processing module is used for continuously collecting and preprocessing ultrasonic transmission signals of the clean membrane body and the membrane bodies with different pollution degrees, extracting ultrasonic transmission signal characteristics under different pollution degrees, and establishing a pollution thickness-signal characteristic mapping model; wherein the ultrasonic transmission signal features comprise a time-frequency domain energy gravity center feature and a frequency attenuation spectrum slope feature; and the pollution detection module is used for judging the pollution stage of the membrane body to be detected based on the pollution thickness-signal feature mapping model, and completing in-situ nondestructive membrane pollution detection. According to the technical scheme, the sludge pollution state in the ceramic membrane and flat membrane filtering process can be monitored in real time, and early warning and pollution dynamic tracking are achieved.

Description

Technical Field

[0001] This invention belongs to the field of wastewater treatment and membrane separation technology, specifically relating to an in-situ non-destructive membrane fouling detection system and method based on ultrasonic transmission signals. Background Technology

[0002] Membrane separation technologies (such as microfiltration, ultrafiltration, and reverse osmosis) are widely used in wastewater treatment, seawater desalination, and food processing. However, membrane fouling is a key issue restricting their long-term stable operation. Fouling leads to decreased membrane flux, increased energy consumption, more frequent cleaning, and even shortened membrane life. Therefore, real-time, accurate, and non-destructive membrane fouling monitoring technology is crucial.

[0003] Currently, membrane fouling detection mainly relies on traditional offline detection methods and online indirect monitoring technologies. Traditional offline detection methods include membrane flux monitoring and offline sampling analysis. The former reflects the overall fouling situation by measuring transmembrane pressure and permeate flow rate, but it has a time lag. The latter uses scanning electron microscopy or infrared spectroscopy to analyze pollutants, but this requires destroying samples and cannot provide real-time monitoring. Online indirect monitoring technologies mainly use optical sensors or electrical impedance spectroscopy. Optical methods are easily affected by water quality and cannot detect internal membrane fouling, while electrical impedance spectroscopy is only sensitive to conductive pollutants. Although ultrasonic detection technology has the advantage of non-destructive penetration, existing schemes are mostly limited to single-frequency detection, focusing only on sound velocity or amplitude attenuation while ignoring envelope energy changes, and lack dynamic fouling models, making it difficult to achieve early warning and fouling type differentiation. These methods have significant shortcomings in terms of real-time performance, sensitivity, and applicability, and cannot meet the urgent needs of wastewater treatment, food processing, and other fields for accurate membrane fouling monitoring. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides an in-situ non-destructive membrane fouling detection system and method based on ultrasonic transmission signals. Its core lies in the construction of a complete dynamic monitoring system, which achieves real-time and accurate monitoring of membrane fouling status through multi-dimensional signal analysis and processing.

[0005] To achieve the above objectives, the present invention provides the following solution: An in-situ non-destructive membrane contamination detection system based on ultrasonic transmission signals includes: An ultrasonic excitation module is used to generate a sinusoidal excitation signal, which penetrates the membrane to excite an ultrasonic transmission signal for contamination detection; The signal processing module is used to continuously acquire and preprocess ultrasonic transmission signals from clean membranes and membranes with different levels of contamination, extract ultrasonic transmission signal features under different levels of contamination, and establish a contamination thickness-signal feature mapping model; wherein, the ultrasonic transmission signal features include time-frequency domain energy centroid features and frequency attenuation spectrum slope features. The contamination detection module is used to determine the contamination stage of the membrane under test based on the contamination thickness-signal feature mapping model, and to complete in-situ non-destructive membrane contamination detection.

[0006] Preferably, the ultrasonic excitation module includes: A function generator is used to generate a sinusoidal excitation signal with a preset frequency and amplitude. The excitation end piezoelectric transducer is used to transmit a sinusoidal excitation signal; The receiving end piezoelectric transducer is used to receive the ultrasonic transmission signal generated after the sinusoidal excitation signal penetrates the membrane. The transmitting transducer and the receiving transducer are symmetrically arranged on both sides of the membrane, aligned and with the same acoustic path.

[0007] Preferably, the signal processing module includes: The signal preprocessing unit is used to denoise the ultrasonic transmission signal using a programmable analog bandpass filter to obtain the filtered ultrasonic transmission signal. The time-frequency domain energy centroid calculation unit is used to perform continuous wavelet transform on the filtered ultrasonic transmission signal to obtain the time spectrum, and calculate the time centroid and frequency centroid of the ultrasonic transmission signal energy based on the time spectrum. The frequency attenuation spectrum slope feature extraction unit is used to perform Hilbert transform on the filtered ultrasonic transmission signal to obtain the envelope signal, calculate the envelope energy based on the envelope signal, and extract the frequency attenuation spectrum slope of the envelope energy by combining ultrasonic pulses with different center frequencies. The mapping model construction unit is used to establish a pollution thickness-signal feature mapping model based on the time centroid and frequency centroid of the ultrasonic transmission signal energy, as well as the slope of the frequency attenuation spectrum of the envelope energy.

[0008] Preferably, the time-frequency domain energy centroid calculation unit includes: A time-frequency energy density calculation subunit is used to calculate the time-frequency energy density based on the time spectrum; wherein, the time spectrum is a time-frequency energy distribution based on a scale factor and a time shift factor; The time centroid calculation subunit is used to calculate the frequency-weighted average time at each time point of the time spectrum within a preset frequency band based on the time-frequency energy density and the time shift factor, and obtain the time centroid; wherein, the delay of the time centroid is used to characterize the increase in the propagation time of the ultrasonic transmission signal in the membrane contamination layer; The frequency centroid calculation subunit is used to calculate the frequency-weighted average frequency of the time spectrum within a preset frequency band based on the time-frequency energy density and the center frequency corresponding to the scale factor, and obtain the frequency centroid; wherein, the downward shift of the frequency centroid is used to characterize the selective absorption of high-frequency energy in the ultrasonic transmission signal by the polluting medium.

[0009] Preferably, the frequency attenuation spectrum slope feature extraction unit includes: The envelope energy calculation subunit is used to calculate the instantaneous energy of the envelope signal based on a preset time window, and to integrate the instantaneous energy within the preset time window to obtain the envelope energy. The attenuation calculation subunit is used to extract the attenuation of each ultrasonic transmission signal relative to a reference signal based on ultrasonic pulses of different center frequencies emitted sequentially; wherein the reference signal is the ultrasonic transmission signal that penetrates a clean membrane. The slope calculation subunit is used to perform linear fitting on different center frequencies and their corresponding attenuation amounts to obtain the slope of the frequency attenuation spectrum.

[0010] This invention also provides an in-situ non-destructive membrane contamination detection method based on ultrasonic transmission signals, wherein the system includes: A sinusoidal excitation signal is used to penetrate the membrane to excite an ultrasonic transmission signal for contamination detection; Ultrasonic transmission signals from clean membranes and membranes with different levels of contamination were continuously acquired and preprocessed. Ultrasonic transmission signal features under different levels of contamination were extracted, and a contamination thickness-signal feature mapping model was established. The ultrasonic transmission signal features include time-frequency domain energy centroid features and frequency attenuation spectrum slope features. Based on the contamination thickness-signal feature mapping model, the contamination stage of the membrane to be tested is determined, and in-situ non-destructive membrane contamination detection is completed.

[0011] Preferred methods for establishing a contamination thickness-signal feature mapping model include: The ultrasonic transmission signal is denoised using a programmable analog bandpass filter to obtain a filtered ultrasonic transmission signal. The time spectrum is obtained by performing continuous wavelet transform on the filtered ultrasonic transmission signal, and the time centroid and frequency centroid of the ultrasonic transmission signal energy are calculated based on the time spectrum. The Hilbert transform of the filtered ultrasonic transmission signal is performed to obtain the envelope signal, and the envelope energy is calculated based on the envelope signal. The frequency attenuation spectrum slope of the envelope energy is extracted by combining ultrasonic pulses with different center frequencies. Based on the time and frequency centroids of the ultrasonic transmission signal energy, as well as the slope of the frequency attenuation spectrum of the envelope energy, a pollution thickness-signal feature mapping model is established.

[0012] Preferably, the method for calculating the time centroid and frequency centroid of the ultrasonic transmitted signal energy includes: Based on the time spectrum, the time-frequency energy density is calculated; wherein, the time spectrum is the time-frequency energy distribution based on the scale factor and the time shift factor; Based on the time-frequency energy density and the time shift factor, the frequency-weighted average time at each time point of the time spectrum within the preset frequency band is calculated to obtain the time centroid; wherein, the delay of the time centroid is used to characterize the increase in the propagation time of the ultrasonic transmission signal in the membrane contamination layer. Based on the time-frequency energy density and the center frequency corresponding to the scale factor, the frequency-weighted average frequency of the time spectrum within the preset frequency band is calculated to obtain the frequency centroid; wherein, the downward shift of the frequency centroid is used to characterize the selective absorption of high-frequency energy in the ultrasonic transmission signal by the polluting medium.

[0013] Compared with existing technologies, the beneficial effects of this invention are as follows: By integrating multi-band ultrasonic excitation and time-domain envelope energy analysis technology, it achieves highly sensitive identification of initial contamination; by innovatively combining time-domain energy decay characteristics and frequency-domain energy distribution patterns, it successfully distinguishes different types of contamination such as particulate deposition and biofilms; and by employing the universally applicable ultrasonic transmission principle, it breaks through the dependence of traditional optical detection on transparent materials, making contamination monitoring of various membrane materials such as ceramic membranes, polymer membranes, and metal membranes possible. This technological breakthrough not only solves the problem that existing methods cannot achieve dynamic tracking of contamination, but also provides a scientific basis for intelligent cleaning decisions in membrane processes by establishing a time-energy decay curve prediction model, filling the technological gap in the field of precise and intelligent membrane contamination monitoring in the industry. Attached Figure Description

[0014] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the in-situ non-destructive membrane contamination detection system based on ultrasonic transmission signals according to an embodiment of the present invention; Figure 2 This is a schematic diagram showing the arrangement of the ultrasonic excitation module according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the filtered signal according to an embodiment of the present invention; Figure 4 This is a schematic diagram of envelope analysis after Hilbert transform in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the change of the time centroid TC with thickness in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the variation of the centroid FC of frequency with thickness in an embodiment of the present invention; Figure 7 This is a schematic diagram illustrating the variation of the attenuation slope S with thickness in an embodiment of the present invention; Figure 8 The envelope energy E in this embodiment of the invention env Schematic diagram showing the variation with thickness. Detailed Implementation

[0016] 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.

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Example 1: like Figure 1 The present invention discloses an in-situ non-destructive membrane contamination detection system based on ultrasonic transmission signals, comprising: an ultrasonic excitation module, a signal processing module, and a contamination detection module.

[0019] An ultrasonic excitation module is used to generate a sinusoidal excitation signal, which penetrates the membrane to excite an ultrasonic transmission signal used for contamination detection. A further embodiment includes: The function generator is used to generate a sinusoidal excitation signal with a preset frequency and amplitude; specifically, the signal is a sinusoidal pulse train modulated by a Hanning window with a finite number of periods, the number of periods being adjustable (1-10 periods) to control the pulse duration and frequency bandwidth. The excitation signal is amplified and applied to the transducer, generating a transmitted wave through the membrane module.

[0020] The excitation end piezoelectric transducer (excitation device) is used to transmit a sinusoidal excitation signal.

[0021] The receiving end piezoelectric transducer (receiving device) is used to receive the ultrasonic transmission signal generated after the sinusoidal excitation signal penetrates the membrane.

[0022] In this configuration, the transmitting and receiving transducers are symmetrically arranged on both sides of the membrane (e.g., a ceramic membrane module), aligned and aligned, and the acoustic path is kept consistent by water. Figure 2 As shown, piezoelectric transducers are symmetrically arranged on both sides of the ceramic membrane or flat sheet membrane assembly, ensuring alignment of the excitation and receiving devices. The hardware components mainly include: a signal acquisition card (ZKCX-YWWSMV1.0) and a water immersion ultrasonic probe (500KHz25Z-SJ). The host computer system represents the signal processing module and the contamination detection module.

[0023] The signal processing module continuously acquires and preprocesses ultrasonic transmission signals from clean membranes and membranes with varying degrees of contamination, extracts ultrasonic transmission signal characteristics under different contamination levels, and establishes a contamination thickness-signal feature mapping model. These ultrasonic transmission signal characteristics include time-frequency domain energy centroid features and frequency attenuation spectrum slope features. After ultrasonic waves penetrate the membrane, surface contamination (such as particle deposition or biofilm formation) causes wave attenuation, scattering, and phase distortion, resulting in changes in the amplitude, envelope shape, and spectral distribution of the transmitted signal.

[0024] A further implementation method is that the signal processing module includes: a signal preprocessing unit, a time-frequency domain energy centroid calculation unit, a frequency attenuation spectrum slope feature extraction unit, and a mapping model construction unit.

[0025] The signal preprocessing unit is used to denoise the ultrasonic transmission signal using a programmable analog bandpass filter to obtain the filtered ultrasonic transmission signal. Specifically, a high-speed analog-to-digital converter acquisition module is used with a sampling rate of 100 MHz and a quantization resolution of 10 bits, which can reproduce the details of the ultrasonic transmission signal with high fidelity. The acquired signal is preprocessed by a programmable analog bandpass filter (passband 100 kHz–1 MHz) to eliminate low-frequency mechanical noise and high-frequency interference. Figure 3 The diagram illustrates the ultrasonic signal reception and preprocessing process of this invention. The piezoelectric transducer at the receiving end converts the transmitted ultrasonic signal into an electrical signal. After signal acquisition, the signal processing module performs the following operations (MATLAB): a bandpass filter is used to eliminate background noise; the upper and lower cutoff frequencies of the bandpass filter are adjustable within the range of 0.5-5 MHz.

[0026] The time-frequency domain energy centroid calculation unit is used to calculate the filtered ultrasonic transmission signal. x(t) Obtaining the time spectrum by performing continuous wavelet transform And based on the time spectrum Calculate the time centroid of the ultrasonic transmitted signal energy respectively With frequency centroid A further implementation method is that the time-frequency domain energy centroid calculation unit includes: Time-frequency energy density calculation subunit, used for time-frequency spectrum calculation Calculate time-frequency energy density Among them, time spectrum Based on scale factor With time shift factor Time-frequency energy distribution; time-frequency energy density The calculation formula is as follows: .

[0027] Time centroid calculation subunit, used for time-frequency energy density calculation and time shift factor Calculate the frequency-weighted average time of the time spectrum at each time point within the preset frequency band to obtain the time centroid. Among them, the center of gravity of time The time delay is used to characterize the increase in the propagation time of the ultrasonic transmitted signal through the contaminated membrane layer; this reflects the concentration of energy on the time axis. Thicker sludge layers lead to a delay in sound wave propagation time and waveform broadening, causing a shift in the time centroid. The calculation formula is as follows: .

[0028] Frequency centroid calculation subunit, used for time-frequency energy density calculation and scale factor corresponding center frequency Calculate the frequency-weighted average frequency of the time spectrum within the preset frequency band to obtain the frequency centroid. Among them, the frequency centroid The downward shift of the frequency centroid is used to characterize the selective absorption of high-frequency energy in ultrasonic transmission signals by the contaminated medium. Sludge exhibits stronger absorption of high-frequency components, and the frequency centroid shifts towards lower frequencies with increasing thickness. (Frequency centroid) The calculation formula is as follows: .

[0029] The physical meanings of the time centroid and frequency centroid are respectively the "weighted average position" of the ultrasonic transmitted signal energy in the time domain and frequency domain, which can intuitively reflect the influence of the fouling layer on the ultrasonic wave propagation path and absorption characteristics. When membrane fouling (especially sludge-type fouling) thickens or its physical state changes, the internal structure of the transmitted signal will produce nonlinear distortion, resulting in a characteristic shift in the energy centroid.

[0030] Specifically, the delay in the center of gravity directly characterizes the increase in the propagation time of ultrasound waves within the contaminant layer. Since the propagation speed of sound waves in sludge media is typically lower than in clean water, the thickening of the contaminant layer is equivalent to lengthening the low-speed propagation path, thereby causing a shift in the entire signal packet along the time axis, and thus increasing the TC value. This change is particularly sensitive to the cumulative thickness of the contaminant layer.

[0031] Meanwhile, the downward shift in the frequency centroid primarily reveals the selective absorption of high-frequency components by the polluting medium. The absorption attenuation coefficient of sound waves by viscoelastic media such as sludge is typically positively correlated with frequency, meaning that high-frequency components attenuate faster than low-frequency components. As the degree of pollution intensifies, a large amount of high-frequency energy in the transmitted signal is consumed, with the remaining energy concentrated in the low-frequency range, resulting in a decrease in the FC value towards lower frequencies. This phenomenon is closely related to the density, viscoelasticity, and other physical properties of the pollutant layer.

[0032] The frequency attenuation spectrum slope feature extraction unit is used to extract features from the filtered ultrasonic transmission signal. x(t) The envelope signal is obtained by performing Hilbert transform, and the envelope energy is calculated based on the envelope signal. The frequency attenuation spectrum slope of the envelope energy is extracted by combining ultrasonic pulses with different center frequencies. A further implementation method includes a frequency attenuation spectrum slope feature extraction unit comprising: The envelope energy calculation subunit is used to calculate the instantaneous energy of the envelope signal based on a preset time window, and integrate the instantaneous energy within the preset time window to obtain the envelope energy; specifically, the analytical signal expression is: , in, x(t) This is the filtered ultrasonic transmission signal. This is the Hilbert transform of the ultrasonic transmission signal. This represents the analytic signal, used to extract the signal envelope. j It represents the imaginary unit.

[0033] Envelope signal for: , Select a time window [t1, t2] in the time domain and calculate the instantaneous energy of the envelope signal. : , The envelope energy is obtained by integrating within the window. : .

[0034] Attenuation calculation subunit, used to calculate different center frequencies emitted sequentially. The ultrasonic pulses are used to extract the attenuation of each transmitted ultrasonic signal relative to the reference signal. The reference signal is the ultrasonic transmission signal that penetrates a clean membrane; specifically, the attenuation can be expressed in logarithmic form of the envelope energy ratio. .

[0035] in, Represents the corresponding center frequency The envelope energy of the ultrasonic transmission signal of the contaminated membrane. Represents the corresponding center frequency The envelope energy of the ultrasonic transmission signal of a clean membrane.

[0036] The slope calculation subunit is used to perform linear fitting on different center frequencies and their corresponding attenuation amounts to obtain the slope of the frequency attenuation spectrum. Specifically, the linear fitting calculates the slope by taking the center frequency... and its corresponding attenuation Plot the points on a coordinate system. Since the attenuation coefficient of sound waves in silt is usually approximately proportional to the frequency, these points will exhibit an approximately linear relationship. Perform linear regression on these data points to obtain the slope. .

[0037] .

[0038] in, This represents the intercept obtained after linear fitting.

[0039] The magnitude of this slope value is directly determined by the physical composition and structure of the silt, making it a sensitive indicator for distinguishing different types of silt. A large slope value indicates that the attenuation increases sharply with increasing frequency. This usually means that the silt medium is highly viscous, has fine particles, and is rich in organic matter or colloidal components. Such media have a strong scattering and absorption effect on sound waves, especially high-frequency sound waves, leading to the rapid loss of high-frequency components. A small slope value indicates that the attenuation is not sensitive to changes in frequency. This may correspond to sediments with coarser particles, higher sand content, and relatively loose structure, whose sound energy loss mechanism is different and less dependent on frequency.

[0040] like Figure 4 The diagram illustrates the signal analysis and feature extraction process of this invention. Time-domain analysis: A Hilbert transform is performed on the signal to extract the envelope curve. The energy metric is obtained by calculating the area of ​​this envelope curve over a specific time interval. This area is calculated using a numerical integration method, preferably the trapezoidal rule, to ensure accuracy and feasibility. The integration interval is set as the effective duration of the ultrasonic transmission signal, i.e., from the point where the signal envelope first exceeds the baseline noise threshold (5% of the maximum amplitude) to the point where it falls back below that threshold.

[0041] The mapping model construction unit is used to establish a pollution thickness-signal feature mapping model based on the time centroid and frequency centroid of the ultrasonic transmission signal energy, as well as the slope of the frequency attenuation spectrum of the envelope energy.

[0042] Specifically, in practical applications, characteristic data under different pollution states can be obtained through calibration experiments to establish a "pollution thickness-signal feature" mapping model. By combining the time-frequency domain energy centroid characteristics and the frequency attenuation spectrum slope characteristics, a feature vector is constructed. F : .

[0043] A pollution characterization matrix with dual parameters in the time and frequency domains is formed for subsequent pollution state identification or thickness prediction.

[0044] The contamination detection module is used to determine the contamination stage of the membrane under test based on the contamination thickness-signal feature mapping model, and to complete in-situ non-destructive membrane contamination detection.

[0045] This embodiment also provides a data processing terminal for performing various functions such as signal processing and pollution detection modules. Based on the aforementioned characteristic data (time-frequency domain energy centroid characteristics and frequency attenuation spectrum slope characteristics), the data processing terminal supports batch analysis and extraction of multiple files. Its main functions include: 1) Import multiple transmission signal files and automatically identify sampling parameters.

[0046] 2) Display the time-domain waveform and the envelope after Hilbert transform for comparison.

[0047] 3) Calculate the envelope energy E of each signal group. env Time-frequency domain energy centroid characteristics TC, FC, and frequency attenuation spectrum slope S.

[0048] 4) Plot the characteristic decay curve, with the horizontal axis representing time or file number and the vertical axis representing the characteristic value. The slope and inflection point of the curve can determine the pollution stage: energy decreases rapidly in the initial stage of pollution, the curve decreases slowly in the middle stage of pollution, and the curve tends to stabilize in the severe stage of pollution. The system can automatically calculate the energy decrease rate and decay coefficient k, realizing real-time pollution trend analysis and early warning.

[0049] Figure 5 , Figure 6 , Figure 7 as well as Figure 8 This invention describes the process for determining the contamination status. A data processing terminal establishes a thickness-feature change curve. In the preliminary experimental phase, a large amount of experimental data is used to verify the relationship between the artificially contaminated test block and the change in the curve's slope. The stage of contamination is determined by the change in the slope. Initial contamination: the curve drops rapidly; intermediate contamination: the slope decreases significantly; severe contamination: the curve tends to flatten.

[0050] This embodiment also provides a visual interface with a modular design, mainly divided into the following functional areas: Real-time monitoring area: displays the time-domain waveform and frequency-domain spectrum of the ultrasonic signal in real time, and plots the time-energy decay curve after fixed intervals; Pollution status area: displays the current pollution level, early warning information, and suggested measures; Parameter setting area: provides an interactive interface for users to adjust detection parameters; Historical data area: supports querying and exporting historical monitoring records. The real-time monitoring area displays the time-domain waveform, frequency-domain spectrum, and characteristic curves. Users can adjust the excitation frequency or sampling rate through the parameter setting area.

[0051] Example 2: This invention also provides an in-situ non-destructive membrane contamination detection method based on ultrasonic transmission signals, and an application system comprising: A sinusoidal excitation signal is used to penetrate the membrane to generate an ultrasonic transmission signal for contamination detection.

[0052] Ultrasonic transmission signals from clean membranes and membranes with different levels of contamination were continuously acquired and preprocessed. Ultrasonic transmission signal features under different levels of contamination were extracted, and a contamination thickness-signal feature mapping model was established. Among them, the ultrasonic transmission signal features include time-frequency domain energy centroid features and frequency attenuation spectrum slope features.

[0053] Based on the contamination thickness-signal feature mapping model, the contamination stage of the membrane to be tested is determined, and in-situ non-destructive membrane contamination detection is completed.

[0054] A further implementation method involves establishing a contamination thickness-signal feature mapping model, including: A programmable analog bandpass filter is used to denoise the ultrasonic transmission signal to obtain the filtered ultrasonic transmission signal.

[0055] The time spectrum is obtained by performing continuous wavelet transform on the filtered ultrasonic transmission signal, and the time centroid and frequency centroid of the ultrasonic transmission signal energy are calculated based on the time spectrum.

[0056] The Hilbert transform of the filtered ultrasonic transmission signal is used to obtain the envelope signal, and the envelope energy is calculated based on the envelope signal. The frequency attenuation spectrum slope of the envelope energy is extracted by combining ultrasonic pulses with different center frequencies.

[0057] Based on the time and frequency centroids of the ultrasonic transmission signal energy, as well as the slope of the frequency attenuation spectrum of the envelope energy, a pollution thickness-signal feature mapping model is established.

[0058] A further implementation method includes a method for calculating the time centroid and frequency centroid of the ultrasonic transmitted signal energy, comprising: The time-frequency energy density is calculated based on the time spectrum; where the time spectrum is the time-frequency energy distribution based on the scale factor and the time shift factor.

[0059] Based on the time-frequency energy density and the time shift factor, the frequency-weighted average time of the time spectrum at each time point within the preset frequency band is calculated to obtain the time centroid; where the delay of the time centroid is used to characterize the increase in the propagation time of the ultrasonic transmission signal in the membrane contamination layer.

[0060] Based on the time-frequency energy density and the center frequency corresponding to the scale factor, the frequency-weighted average frequency of the time spectrum within the preset frequency band is calculated to obtain the frequency centroid; wherein, the downward shift of the frequency centroid is used to characterize the selective absorption of high-frequency energy in the ultrasonic transmission signal by the polluting medium.

[0061] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. An in-situ non-invasive membrane fouling detection system based on ultrasonic transmission signals, characterized in that, The method comprises the following steps: An ultrasonic excitation module is used to generate a sinusoidal excitation signal that penetrates the membrane to excite an ultrasonic transmission signal for contamination detection; A signal processing module is used to continuously collect and pre-process the ultrasonic transmission signals of clean membranes and membranes with different degrees of contamination, extract ultrasonic transmission signal features under different degrees of contamination, and establish a contamination thickness-signal feature mapping model; wherein the ultrasonic transmission signal features include time-frequency energy center features and frequency attenuation spectrum slope features; A contamination detection module is used to determine the contamination stage of the membrane to be detected based on the contamination thickness-signal feature mapping model, and complete in-situ non-destructive membrane contamination detection.

2. The system of claim 1, wherein, The ultrasonic excitation module comprises: A function generator is used to generate a sinusoidal excitation signal with a preset frequency and amplitude; An excitation end piezoelectric transducer is used to emit the sinusoidal excitation signal; A receiving end piezoelectric transducer is used to receive the ultrasonic transmission signal generated after the sinusoidal excitation signal penetrates the membrane; Wherein, the transmitting end transducer and the receiving end transducer are symmetrically arranged on both sides of the membrane, aligned and have consistent sound paths.

3. The system of claim 1, wherein, The signal processing module comprises: A signal preprocessing unit is used to denoise the ultrasonic transmission signal using a programmable analog bandpass filter to obtain a filtered ultrasonic transmission signal; A time-frequency energy center calculation unit is used to perform continuous wavelet transform on the filtered ultrasonic transmission signal to obtain a time-frequency spectrum, and calculate the time center and frequency center of the ultrasonic transmission signal energy based on the time-frequency spectrum; A frequency attenuation spectrum slope feature extraction unit is used to perform Hilbert transform on the filtered ultrasonic transmission signal to obtain an envelope signal, calculate the envelope energy based on the envelope signal, and extract the frequency attenuation spectrum slope of the envelope energy in combination with ultrasonic pulses of different center frequencies; A mapping model construction unit is used to establish a contamination thickness-signal feature mapping model based on the time center and frequency center of the ultrasonic transmission signal energy and the frequency attenuation spectrum slope of the envelope energy.

4. The system of claim 3, wherein, The time-frequency energy center calculation unit comprises: A time-frequency energy density calculation subunit is used to calculate the time-frequency energy density based on the time-frequency spectrum; wherein the time-frequency spectrum is a time-frequency energy distribution based on a scale factor and a time shift factor; A time center calculation subunit is used to calculate the frequency-weighted average time at each time point of the time-frequency spectrum within a preset frequency band based on the time-frequency energy density and the time shift factor to obtain the time center; wherein the delay of the time center is used to represent the increase in the propagation time of the ultrasonic transmission signal in the membrane contamination layer; A frequency center calculation subunit is used to calculate the frequency-weighted average frequency of the time-frequency spectrum within a preset frequency band based on the time-frequency energy density and the center frequency corresponding to the scale factor to obtain the frequency center; wherein the downward movement of the frequency center is used to represent the selective absorption of high-frequency energy in the ultrasonic transmission signal by the contamination medium.

5. The system of claim 3, wherein, The frequency attenuation spectrum slope feature extraction unit comprises: An envelope energy calculation subunit is used to calculate the instantaneous energy of the envelope signal based on a preset time window, and integrate the instantaneous energy within the preset time window to obtain the envelope energy; The attenuation amount calculation subunit is configured to extract an attenuation amount of each ultrasonic transmission signal relative to a reference signal based on the ultrasonic pulses with different center frequencies transmitted in sequence, wherein the reference signal is an ultrasonic transmission signal penetrating a clean membrane body; The slope calculation subunit is configured to perform linear fitting on the different center frequencies and the corresponding attenuation amounts to obtain a frequency attenuation spectrum slope.

6. An in-situ non-invasive membrane fouling detection method based on ultrasonic transmission signals, applying the system of any one of claims 1-5, characterized in that, The method comprises: penetrating the membrane body with a sine wave excitation signal to excite an ultrasonic transmission signal for contamination detection; continuously collecting and preprocessing ultrasonic transmission signals of the clean membrane body and the membrane bodies with different contamination levels, extracting ultrasonic transmission signal features under different contamination levels, and establishing a contamination thickness-signal feature mapping model; wherein the ultrasonic transmission signal features include time-frequency energy barycenter features and frequency attenuation spectrum slope features; based on the contamination thickness-signal feature mapping model, determining the contamination stage of the membrane body to be detected to complete in-situ non-destructive membrane contamination detection.

7. The method of claim 6, wherein, The method for establishing the contamination thickness-signal feature mapping model comprises: using a programmable analog band-pass filter to denoise the ultrasonic transmission signal to obtain a filtered ultrasonic transmission signal; performing continuous wavelet transform on the filtered ultrasonic transmission signal to obtain a time-frequency spectrum, and calculating the time barycenter and the frequency barycenter of the ultrasonic transmission signal energy based on the time-frequency spectrum; performing Hilbert transform on the filtered ultrasonic transmission signal to obtain an envelope signal, and calculating envelope energy based on the envelope signal, and combining the ultrasonic pulses with different center frequencies to extract the frequency attenuation spectrum slope of the envelope energy; based on the time barycenter and the frequency barycenter of the ultrasonic transmission signal energy and the frequency attenuation spectrum slope of the envelope energy, establishing the contamination thickness-signal feature mapping model.

8. The method of claim 7, wherein, The method for calculating the time barycenter and the frequency barycenter of the ultrasonic transmission signal energy comprises: based on the time-frequency spectrum, calculating a time-frequency energy density; wherein the time-frequency spectrum is a time-frequency energy distribution based on a scale factor and a time shift factor; based on the time-frequency energy density and the time shift factor, calculating a frequency-weighted average time of each time point on the time-frequency spectrum in a preset frequency band to obtain the time barycenter; wherein the delay of the time barycenter is used to represent the increase in the propagation time of the ultrasonic transmission signal in the membrane contamination layer; based on the time-frequency energy density and the center frequency corresponding to the scale factor, calculating a frequency-weighted average frequency of the time-frequency spectrum in the preset frequency band to obtain the frequency barycenter; wherein the downward shift of the frequency barycenter is used to represent the selective absorption of the high-frequency energy of the ultrasonic transmission signal by the contamination medium.

Images