Audio-based intelligent monitoring system and method for pipeline leakage

Abstract

The present application relates to the technical field of pipeline safety monitoring, and specifically relates to a pipeline leakage intelligent monitoring system and method based on audio frequency, comprising an audio signal acquisition module, a noise filtering and feature extraction module, a leakage identification and positioning module, and a state evaluation and alarm module; wherein: the audio signal acquisition module acquires pipeline audio signals and outputs original audio signal data; the noise filtering and feature extraction module performs filtering processing to suppress environmental noise and extracts a feature vector related to leakage; the leakage identification and positioning module identifies whether a leakage event exists and calculates the position of a leakage point relative to adjacent sensors; and the state evaluation and alarm module generates a leakage state evaluation result. Through the construction of a closed-loop monitoring process from audio frequency acquisition to alarm output, the present application realizes accurate identification, spatial positioning and hierarchical evaluation of pipeline leakage events, and significantly improves the response efficiency and practicality of the monitoring system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipeline safety monitoring, and in particular to a pipeline leakage intelligent monitoring system and method based on acoustic frequency. BACKGROUND

[0002] In the pipeline systems of urban water supply, oil transportation, natural gas transmission and chemical processes, once leakage occurs, it will lead to energy waste, environmental pollution and even safety accidents, causing serious economic and social losses. Traditional pipeline leakage detection methods rely on manual inspection, pressure monitoring or flow comparison, but such methods are limited by high response delay, low spatial positioning accuracy and the inability to reflect the local leakage state in real time, and are difficult to meet the actual needs of modern complex pipeline networks for high-precision, rapid response and remote integrated monitoring. In recent years, acoustic frequency monitoring technology has attracted widespread attention in the field of pipeline leakage detection due to its advantages of non-contact, high sensitivity and real-time deployment.

[0003] However, existing acoustic frequency-based monitoring solutions mostly only stay at the level of single-point feature analysis or rough energy judgment, and fail to achieve the unified integration of multi-dimensional feature extraction, event level evaluation and precise spatial positioning of leakage signals. At the same time, the processing steps of the system are often scattered and isolated, lacking a closed-loop automated processing mechanism from acoustic frequency acquisition to alarm output, resulting in low accuracy of leakage detection and inability to meet the needs of rapid and reliable response in industrial applications. Therefore, there is an urgent need for an acoustic frequency-based pipeline leakage intelligent monitoring system and method to improve the timeliness and effectiveness of leakage identification. SUMMARY

[0004] Based on the above purpose, the present application provides an acoustic frequency-based pipeline leakage intelligent monitoring system and method.

[0005] The acoustic frequency-based pipeline leakage intelligent monitoring system comprises an acoustic frequency signal acquisition module, a noise filtering and feature extraction module, a leakage identification and positioning module, and a state evaluation and alarm module. Wherein:

[0006] The acoustic frequency signal acquisition module is used to acquire pipeline acoustic frequency signals through acoustic sensors arranged along the pipeline, and outputs raw acoustic frequency signal data.

[0007] The noise filtering and feature extraction module is used to receive the raw acoustic frequency signal data, filter it to suppress environmental noise, and extract a feature vector related to leakage.

[0008] The leakage identification and positioning module is used to receive the feature vector, identify whether there is a leakage event, and calculate the position of the leakage point relative to the adjacent sensors based on the signal strength and propagation attenuation model after confirming the leakage, and output the leakage identification result and position information.

[0009] The state evaluation and alarm module is configured to generate a leakage state evaluation result according to the identification result of the leakage event and the position information thereof, and output an alarm signal to an upper computer or a management terminal.

[0010] Optionally, the audio signal acquisition module comprises a sensor arrangement unit, a signal sampling unit and a data output unit, wherein:

[0011] The sensor arrangement unit is configured to arrange a plurality of acoustic sensors along the pipeline at a preset interval, and attach the sensors to the outer side of the pipeline wall, and record the installation position coordinates of the sensors along the pipeline through positioning tags;

[0012] The signal sampling unit is configured to control the acoustic sensors to synchronously collect the acoustic wave vibration of the inner wall of the pipeline at a fixed sampling frequency, and generate a continuous time-domain audio signal sequence;

[0013] The data output unit is configured to mark the audio signal sequence collected by each sensor with a time stamp, package the audio signal sequence in a unified format, and output the original audio signal data.

[0014] Optionally, the noise filtering and feature extraction module comprises a frequency band filtering unit, a time-frequency transformation unit and a feature extraction unit, wherein:

[0015] The frequency band filtering unit is configured to receive the original audio signal data output by the audio signal acquisition module, and filter the signal in the frequency domain by using a band-pass filter to filter out low-frequency mechanical vibration noise and high-frequency electromagnetic interference in the background environment, and retain the leakage-related audio components in the preset frequency band;

[0016] The time-frequency transformation unit is configured to perform short-time Fourier transform processing on the filtered audio signal to generate a time-frequency distribution map of the audio signal;

[0017] The feature extraction unit is configured to extract the mutation point position, high-frequency continuous segment energy value and spectral slope feature related to the leakage in the time-frequency map, and output the features in the form of a feature vector.

[0018] Optionally, the feature extraction unit comprises:

[0019] The mutation point detection subunit is configured to extract high-energy mutation points along the time axis in the time-frequency map, calculate the spectral energy difference sequence of adjacent time frames, and determine the energy mutation position based on a set mutation threshold .

[0020] The high-frequency band identification subunit is configured to identify the high-frequency band signal that continuously appears in the time-frequency map, set a frequency threshold , calculate the average energy greater than , and screen the continuous high-frequency band segment with a duration not less than a set length .

[0021] Spectrum envelope analysis subunit: for extracting the trend of the change of the spectrum envelope in the continuous frame, constructing an envelope change curve by normalizing the envelope energy, and calculating the change rate by using the first derivative to output the spectrum slope feature;

[0022] Feature output: for splicing the output of the mutation point position, the high frequency segment energy value and the spectrum slope feature to form a fixed dimension feature vector.

[0023] Optionally, the leakage identification and positioning module comprises a leakage event identification unit, a signal strength analysis unit and a leakage positioning calculation unit; wherein:

[0024] Leakage event identification unit: for receiving the feature vector output by the feature extraction unit, sequentially obtaining the high-energy mutation amplitude, the continuous high-frequency segment average energy value and the spectrum slope value in the feature vector, and comparing them with the preset leakage identification threshold value respectively; when any one of the feature values exceeds the corresponding threshold value, a leakage confirmation signal is output;

[0025] Signal strength analysis unit: for receiving the filtered audio signals of each sensor in the audio signal acquisition module, calculating the signal strength value corresponding to each sensor, and recording the corresponding sensor spatial position coordinates;

[0026] Leakage positioning calculation unit: for receiving the leakage confirmation signal, and based on the signal strength values of multiple adjacent sensors and their spatial positions, deriving the relative position coordinates of the leakage point through the propagation attenuation model, and outputting the identification result of the leakage event and the corresponding spatial position information.

[0027] Optionally, the signal strength analysis unit comprises:

[0028] Envelope extraction subunit: for receiving the filtered audio signals of each sensor in the audio signal acquisition module, extracting the instantaneous amplitude envelope thereof through Hilbert transform, and generating the signal envelope sequence of each sensor channel;

[0029] Intensity calculation subunit: for calculating the energy average value of each sensor's signal envelope sequence to obtain the audio signal strength value of the corresponding sensor ;

[0030] Position mapping subunit: for retrieving the preset sensor installation layout information, obtaining the three-dimensional spatial coordinate value corresponding to each sensor, and binding and storing the calculated signal strength value.

[0031] Optionally, the leakage positioning calculation unit comprises:

[0032] The sensor pair screening subunit is configured to, after receiving the leakage confirmation signal, screen out, from all sensors, a plurality of adjacent sensors with signal strength values higher than a set threshold value to form an effective sensing pair, and reserve spatial position coordinates and corresponding strength values of the effective sensing pair;

[0033] The strength difference calculation subunit is configured to calculate a difference between the audio signal strength values of any pair of sensors and , and deduce a distance difference between the leakage point and the two sensors based on an energy attenuation relationship of sound waves in a pipeline medium ;

[0034] The position inversion derivation subunit is configured to, based on a plurality of groups and corresponding spatial position coordinates, derive relative spatial position coordinates of the leakage point by a least square method ;

[0035] The output result is that the spatial coordinates of the leakage point derived by inversion are taken as a positioning result of the leakage event, and are output together with corresponding leakage confirmation information to form a final leakage identification result and spatial position information.

[0036] Optionally, the state evaluation and alarm module comprises a leakage level determination subunit, a state generation subunit and an alarm output subunit; wherein:

[0037] The leakage level determination subunit is configured to calculate a leakage influence range and a strength level according to the leakage identification result and spatial position information output by the leakage identification and positioning module, in combination with corresponding signal strength values and a multi-source sensor coverage situation, and to perform a graded determination on the leakage event according to a preset evaluation standard, and to output a leakage level label;

[0038] The state generation subunit is configured to generate a standardized leakage state evaluation result according to the leakage level label and the leakage position coordinates, including an event number, an occurrence time, a leakage level, spatial coordinates and a current processing state mark;

[0039] The alarm output subunit is configured to, after the leakage state evaluation result is generated, send alarm information to an upper computer or a management terminal through an industrial communication interface.

[0040] Optionally, the leakage level determination subunit comprises:

[0041] The strength index calculation subunit is configured to obtain spatial positions of the leakage points output by the leakage identification and positioning module, and extract signal strength values of a plurality of sensors near the spatial positions, and calculate an average strength index of a leakage area ;

[0042] The influence range estimation subunit is configured to analyze a plurality of sensors that satisfy a condition that a signal strength value is higher than a preset strength threshold value the number of sensors and its distribution space range, and then the leakage influence radius is calculated ;

[0043] Grade decision generation subunit: for matching in the preset leakage grade evaluation standard according to the calculated and value, determining the grade to which the leakage event belongs, and outputting the corresponding leakage grade label .

[0044] The pipeline leakage intelligent monitoring method based on audio is realized by the pipeline leakage intelligent monitoring system based on audio, comprising the following steps:

[0045] S1: multiple acoustic sensors are arranged at preset positions along the pipeline, original audio signal data in the pipeline operation process is collected, and the spatial position information of each sensor is marked;

[0046] S2: the original audio signal obtained in S1 is subjected to band-pass filtering processing to suppress low-frequency mechanical interference and high-frequency electromagnetic noise, then the high-energy mutation point, the continuous high-frequency band average energy value and the spectrum slope value are calculated, and a feature vector is generated;

[0047] S3: the feature vector generated in S2 is compared with the preset leakage identification threshold value in sequence, if any feature dimension exceeds the corresponding threshold value, a leakage confirmation signal is output, and the current detection frame is determined as a leakage state;

[0048] S4: after the leakage state is confirmed, the filtered signals of multiple sensors adjacent to the leakage point in S1 are extracted, the corresponding signal intensity values are calculated, and the spatial coordinate information is called as subsequent positioning input;

[0049] S5: based on the signal intensity values and the spatial positions of the adjacent multiple sensors obtained in S4, a sound wave propagation attenuation model is constructed, the distance difference from the leakage point to each sensor is deduced, and the relative spatial position coordinates of the leakage point are calculated by using the least square method inversion;

[0050] S6: receiving the leakage confirmation signal and the leakage point coordinates output by S3 and S5, combining the leakage average intensity and the influence radius to calculate the leakage grade, packaging the leakage state evaluation result, and outputting the alarm information to the upper computer or the management terminal through the communication interface.

[0051] The beneficial effects of the present application are:

[0052] The present application realizes audio signal acquisition, feature extraction, leakage event identification, spatial positioning and state evaluation in sequence by constructing a complete audio monitoring processing flow, organically combines multi-source sensing information with time-frequency analysis, propagation modeling and grade judgment mechanism, and improves the accuracy and response efficiency of leakage detection.

[0053] The present application can reverse the specific position of the leakage point based on the multi-sensor signal strength difference and the spatial layout after the leakage event occurs, and output structured state information and multi-level alarm signals according to the leakage intensity and the influence range matching the preset grading standard, so as to realize the automatic closed-loop control from perception to response, and effectively improve the real-time monitoring and early warning capability for sudden leakage accidents. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only illustrate the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0055] Figure 1 A schematic diagram of a pipeline leakage intelligent monitoring system according to an embodiment of the present application;

[0056] Figure 2 A schematic diagram of a pipeline leakage intelligent monitoring method according to an embodiment of the present application. DETAILED DESCRIPTION

[0057] The present application will be described in detail below with reference to the drawings and specific embodiments. It should be noted that in order to make the embodiments more detailed, the following embodiments are the best, preferred embodiments, and other alternative ways can also be used by those skilled in the art to implement them; and the drawings are only used to more specifically describe the embodiments, and are not intended to specifically limit the present application.

[0058] It should be noted that in the specification, "one embodiment", "embodiment", "exemplary embodiment", "some embodiments" and the like indicate that the described embodiments can include a specific feature, structure or property, but not necessarily every embodiment includes the specific feature, structure or property. In addition, when a specific feature, structure or property is described in combination with an embodiment, it should be within the knowledge of those skilled in the related art to realize this feature, structure or property in combination with other embodiments (whether or not explicitly described).

[0059] Generally, the terms can be understood at least in part from the usage in context. For example, depending on the context, the term "one or more" used herein can be used to describe any feature, structure or property in a singular sense or in a combination of features, structures or properties in a plural sense. In addition, the term "based on" can be understood as not necessarily conveying a set of exclusive factors, but can instead allow the presence of other factors not necessarily explicitly described, at least in part depending on the context.

[0060] As shown in Figure 1 The audio-based pipeline leakage intelligent monitoring system comprises an audio signal acquisition module, a noise filtering and feature extraction module, a leakage identification and positioning module, and a state evaluation and alarm module; wherein:

[0061] The audio signal acquisition module is used for acquiring pipeline audio signals through acoustic sensors arranged along the pipeline, and outputs raw audio signal data.

[0062] The noise filtering and feature extraction module is used for receiving raw audio signal data, filtering the data to suppress environmental noise, and extracting a feature vector related to leakage.

[0063] The leakage identification and positioning module is used for receiving the feature vector, identifying whether a leakage event exists, and calculating the position of the leakage point relative to adjacent sensors based on signal strength and propagation attenuation models after confirming the leakage, and outputs leakage identification results and position information.

[0064] The state evaluation and alarm module is used for generating a leakage state evaluation result according to the identification result of the leakage event and its position information, and outputting an alarm signal to an upper computer or a management terminal.

[0065] The audio signal acquisition module comprises a sensor arrangement unit, a signal sampling unit, and a data output unit; wherein:

[0066] The sensor arrangement unit is used for arranging a plurality of acoustic sensors along the pipeline at a predetermined interval, with the sensors being installed on the outside of the pipe wall and recording their installation position coordinates along the pipeline through positioning tags for spatial reference in subsequent leakage positioning.

[0067] The signal sampling unit is used for controlling each acoustic sensor to synchronously collect the vibration of the sound wave on the inner wall of the pipeline at a fixed sampling frequency, generating a continuous time-domain audio signal sequence, wherein the sampling frequency is set to be more than twice the maximum frequency component of the leakage audio signal to meet the Nyquist sampling theorem requirement.

[0068] The data output unit is used for time-stamping and uniformly packaging the audio signal sequence collected by each sensor, and outputting raw audio signal data; the above units are structured and refined through the audio signal acquisition module, and the functions of sensor arrangement, sampling, and data output are clearly divided, which can effectively improve the spatial resolution and time synchronization of audio acquisition, and provide high-quality raw signal data support for subsequent feature extraction and leakage positioning.

[0069] The noise filtering and feature extraction module comprises a frequency band filtering unit, a time-frequency transformation unit, and a feature extraction unit; wherein:

[0070] Frequency band filtering unit: Used to receive the raw audio signal data output by the audio signal acquisition module, and use a bandpass filter to perform frequency domain filtering on the signal to filter out low-frequency mechanical vibration noise and high-frequency electromagnetic interference in the background environment, while retaining leakage-related audio components located in the preset frequency band;

[0071] Time-frequency transformation unit: used to perform short-time Fourier transform processing on the filtered audio signal to generate a time-frequency distribution spectrum of the audio signal, reflecting the changes in signal intensity in both time and frequency dimensions;

[0072] Feature extraction unit: used to extract the location of leakage-related abrupt change points, high-frequency duration energy values ​​and spectral slope features from the time-frequency spectrum, and output them as feature vectors as input data for subsequent leakage identification and localization modules; the above unit can effectively suppress external interference noise and enhance the identifiability of leakage feature signals by introducing bandpass filtering and time-frequency transformation mechanisms. The extracted multi-dimensional audio features have stronger stability and discriminative power, thereby improving the accuracy and reliability of subsequent leakage identification.

[0073] The feature extraction unit includes:

[0074] The mutation point detection subunit is used to extract high-energy mutation points along the time axis in the time-frequency spectrum, calculate the spectral energy difference sequence between adjacent time frames, and base the detection on a set mutation threshold. The formula for determining the location of the energy mutation is: ,in, For the first The spectral energy abrupt change value of the frame; For the first Frame at frequency Energy at the location; For the first Frame at frequency Energy at the location; , The minimum and maximum frequency limits are set.

[0075] High-frequency band identification subunit: used to identify high-frequency band signals that continuously appear in the time-frequency spectrum and set the frequency threshold. Statistical greater than The average energy, and screening for those that meet the requirement of a duration not less than a set length. The formula for calculating the average energy of a continuous high-frequency band is: ,in, For the first The average energy of the frame in the high-frequency band; For the first Frame at frequency Energy at the location; Indicates from arrive The number of frequency points between them;

[0076] Spectral envelope analysis subunit: Used to extract the variation trend of the spectral envelope in consecutive frames. It constructs the envelope variation curve by normalizing the envelope energy, calculates its rate of change using the first derivative, and outputs the spectral slope characteristics. The specific calculation formula is as follows: ,in, For the first The first time derivative of the total energy of the frame spectrum; This indicates the operation of differentiation with respect to time;

[0077] Feature output: This feature vector is formed by concatenating the output abrupt change location, high-frequency sustained energy value, and spectral slope features into a fixed-dimensional feature vector, which serves as the input for subsequent identification. Through the above sub-units, the energy abrupt change, sustained high frequency, and spectral change trend are quantified and extracted, and output as a unified feature vector. This can construct a highly discriminative audio feature expression method, enhance the modeling ability of pipeline leakage sound features, and improve the accuracy of subsequent identification and location.

[0078] The leak identification and location module includes a leak event identification unit, a signal strength analysis unit, and a leak location calculation unit; wherein:

[0079] Leakage event identification unit: It is used to receive the feature vector output by the feature extraction unit, sequentially obtain the high energy mutation amplitude, the average energy value of the continuous high frequency band and the spectral slope value in the feature vector, and compare them with the preset leakage identification threshold respectively; when any feature value exceeds the corresponding threshold, a leakage confirmation signal is output.

[0080] Signal strength analysis unit: used to receive the filtered audio signals from each sensor in the audio signal acquisition module, calculate the signal strength value corresponding to each sensor, and record the corresponding spatial position coordinates of the sensor as the input basis for subsequent positioning calculations;

[0081] Leakage location calculation unit: After receiving a leakage confirmation signal, it is used to deduce the relative position coordinates of the leakage point based on the signal strength values ​​of multiple adjacent sensors and their spatial positions through a propagation attenuation model, and output the identification result of the leakage event and its corresponding spatial position information. By combining audio feature recognition with physical model calculation, the above unit can not only accurately determine whether a leakage exists, but also realize the spatial location of the leakage point based on the intensity attenuation law, thereby significantly improving the leakage detection accuracy and practicality of the system.

[0082] The signal strength analysis unit includes:

[0083] Envelope extraction subunit: used for receiving the filtered audio signals of each sensor in the audio signal acquisition module, extracting the instantaneous amplitude envelope thereof through Hilbert transform, and generating the signal envelope sequence of each sensor channel;

[0084] Intensity calculation subunit: used for calculating the energy mean value of the signal envelope sequence of each sensor to obtain the audio signal intensity value of the corresponding sensor , the specific calculation formula is as follows: ; wherein, is the signal intensity value of the i th sensor; is the filtered audio signal envelope of the i th sensor; is the duration of the signal analysis time window;

[0085] Position mapping subunit: used for calling the preset sensor installation layout information, obtaining the three-dimensional space coordinate value corresponding to each sensor, and binding and storing the calculated signal intensity value, for subsequent positioning model calling.

[0086] The step of extracting the instantaneous amplitude envelope thereof through Hilbert transform is as follows:

[0087] First, the filtered time-domain audio signal of the i th sensor is obtained from the audio signal acquisition module, denoted as , and taken as the input signal of Hilbert transform;

[0088] Then, the Hilbert transform is performed on to obtain the corresponding Hilbert transform component , and further construct the analytic signal , the expression of which is: , wherein, is the analytic signal of the i th sensor; is the filtered audio signal; is the orthogonal component obtained by performing Hilbert transform on ; and is the imaginary unit;

[0089] Next, the modulus of the analytic signal is taken to calculate the instantaneous amplitude envelope thereof, that is, the signal envelope sequence is obtained, the expression of which is: ;

[0090] Finally, the calculated is taken as the input of the subsequent intensity calculation subunit for performing energy integration operation to complete signal intensity calculation; ​​​​

[0091] The above step extracts the instantaneous amplitude envelope by Hilbert transform, which can effectively depict the energy fluctuation characteristics of the audio signal in the local time period, avoid the amplitude distortion caused by the superposition of frequency components or short-time phase disturbance, and provide an accurate basis for the robust evaluation of signal strength and subsequent leakage positioning.

[0092] The leakage positioning calculation unit comprises:

[0093] The sensor pair screening subunit is configured to, after receiving the leakage confirmation signal, screen out, from all sensors, a plurality of adjacent sensors with signal strength values higher than a set threshold to form an effective sensing pair, and reserve the spatial position coordinates and corresponding strength values thereof.

[0094] The strength difference calculation subunit is configured to calculate the difference between the audio signal strength values of any pair of sensors and Based on the energy attenuation relationship in the propagation process of the sound wave in the pipeline medium, the distance difference between the leakage point and the two sensors is derived as , and the calculation formula is as follows: , wherein, is the distance difference between the leakage point and the sensor and the sensor . is the signal strength value received by the sensor and . is the attenuation coefficient of the sound wave in the propagation process in the pipeline medium, and is a preset constant.

[0095] The position inversion derivation subunit is configured to, based on a plurality of groups of and corresponding spatial position coordinates, derive the relative spatial position coordinates of the leakage point by least squares method , that is, to solve the point that minimizes the error function, and the error function is defined as: , wherein, : positioning residual error function. is the spatial position vector of the leakage point to be solved. is the spatial coordinate vector of the sensor and . represents the Euclidean distance.

[0096] The output result: the spatial coordinates of the leakage point obtained by inversion are taken as the positioning result of the leakage event, and are output together with the corresponding leakage confirmation information to form the final leakage identification result and spatial position information; the above-mentioned subunit can significantly improve the leakage positioning accuracy by constructing the corresponding relationship between the signal strength difference and the spatial distance difference and combining the joint constraints of multiple pairs of sensors, has strong adaptability and anti-interference ability, and is particularly suitable for the industrial pipeline scene with multiple point layouts and high positioning accuracy requirements.

[0097] The state evaluation and alarm module includes a leakage level determination unit, a state generation unit, and an alarm output unit; wherein:

[0098] The leakage level determination unit: used for calculating the leakage influence range and intensity level according to the leakage identification result and spatial position information output by the leakage identification and positioning module, combining the corresponding signal strength value and the multi-source sensor coverage, and performing graded determination on the leakage event according to the preset evaluation standard, and outputting the leakage level label;

[0099] The state generation unit: used for generating a standardized leakage state evaluation result according to the leakage level label and the leakage position coordinates, including event number, occurrence time, leakage level, spatial coordinates, and current processing state mark;

[0100] The alarm output unit: used for sending alarm information to the upper computer or management terminal through the industrial communication interface after the leakage state evaluation result is generated, supporting Modbus, RS485 or Ethernet TCP protocol transmission format, and setting multi-level response marks according to the alarm level to complete the event triggering reporting; the above-mentioned unit can realize the rapid response and remote management of different severity leakage events by clear grading and standardized state output of the leakage level, and the alarm mechanism combined with the communication protocol adaptation can effectively improve the practicability and deployment flexibility of the system in the industrial environment.

[0101] The leakage level determination unit includes:

[0102] The intensity index calculation subunit: used for obtaining the spatial position of the leakage point output by the leakage identification and positioning module, and extracting the signal strength values of several sensors near the spatial position, and calculating the average intensity index of the leakage area , the calculation formula is: , wherein, is the average intensity index of the leakage area; is the signal strength value of the th sensor within the threshold range from the leakage point; is the number of sensors meeting the spatial threshold condition;

[0103] Impact range estimation subunit: for analyzing the number of sensors satisfying the condition of signal strength value being higher than the preset intensity threshold and its spatial range of distribution, and then calculating the leakage impact radius , the formula of which is: , wherein, is the leakage impact radius; is the spatial position vector of the i-th sensor; is the spatial position vector of the leakage point; is the Euclidean distance; is the preset signal strength determination threshold; represents all sensors satisfying the condition , wherein the index number of the sensor with the largest distance from the leakage point position ;

[0104] Grade decision generation subunit: for matching the calculated and values with the preset leakage grade evaluation standard, determining the grade to which the leakage event belongs, and outputting the corresponding leakage grade label ; the above units can accurately evaluate the actual impact degree of the leakage event by combining signal strength and spatial distribution information, and form clear risk classification labels in cooperation with the preset grade standard, thereby supporting more targeted alarm strategies and subsequent intervention measures.

[0105] Table 1: Leakage grade evaluation mapping table

[0106]

[0107] In the above table 1, the average intensity index is used to measure the overall energy level of the leakage sound source, with the unit of dB; the impact radius represents the spatial scale that can be effectively covered by the leakage sound energy; the grade label represents the discrete grade in the output result, which is used to identify the emergency degree and disposal priority of the leakage event; by establishing the leakage grade evaluation mapping table, the joint quantitative evaluation of the intensity and spatial impact of the leakage event can be realized, making the grade label determination objective, adjustable and real-time executable, and providing clear basis for subsequent alarm grading output and event response.

[0108] As shown in Figure 2 , the pipeline leakage intelligent monitoring method based on sound frequency is realized by the above-mentioned pipeline leakage intelligent monitoring system based on sound frequency, comprising the following steps:

[0109] S1: multiple acoustic sensors are arranged at preset positions along the pipeline, original sound frequency signal data in the pipeline operation process is collected, and the spatial position information of each sensor is marked;

[0110] ​S2: The original audio signal obtained in S1 is subjected to band-pass filtering to suppress low-frequency mechanical interference and high-frequency electromagnetic noise, and then high-energy mutation points, continuous high-frequency band average energy values and spectral slope values are calculated to generate a feature vector;

[0111] S3: The feature vector generated in S2 is compared with preset leakage identification thresholds in sequence, and if any feature dimension exceeds the corresponding threshold, a leakage confirmation signal is output, and the current detection frame is determined as a leakage state;

[0112] S4: After the leakage state is confirmed, a plurality of sensor filtered signals spatially adjacent to the leakage point in S1 are extracted, corresponding signal intensity values are calculated, and spatial coordinate information is called as subsequent positioning input;

[0113] S5: Based on the signal intensity values and spatial positions of the adjacent plurality of sensors obtained in S4, a sound wave propagation attenuation model is constructed, distance differences from the leakage point to each sensor are derived, and the relative spatial position coordinates of the leakage point are calculated by least square inversion;

[0114] S6: The leakage confirmation signal and the leakage point coordinates output by S3 and S5 are received, the leakage level is calculated in combination with the average leakage intensity and the influence radius, the leakage state evaluation result is packaged, and alarm information is output to the upper computer or management terminal through the communication interface.

[0115] The present application encompasses any alternative, modification, equivalent method and scheme made on the essence and scope of the present application. In order to make the public have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can also be fully understood without the description of these details to those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits are not described in detail.

[0116] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be regarded as the protection scope of the present application.

Claims

1. An audio based intelligent monitoring system for pipeline leak detection, characterized in that, The system comprises an audio signal acquisition module, a noise filtering and feature extraction module, a leakage identification and positioning module, and a state evaluation and alarm module. The audio signal acquisition module is configured to acquire pipeline audio signals through acoustic sensors arranged along the pipeline and output raw audio signal data. The noise filtering and feature extraction module is configured to receive the raw audio signal data, filter the data to suppress environmental noise, and extract a feature vector related to leakage. The leakage identification and positioning module is configured to receive the feature vector, identify whether a leakage event exists, and calculate the position of the leakage point relative to adjacent sensors based on signal strength and a propagation attenuation model after confirming the leakage, and output a leakage identification result and position information. The state evaluation and alarm module is configured to generate a leakage state evaluation result based on the identification result of the leakage event and its position information, and output an alarm signal to an upper computer or management terminal. The noise filtering and feature extraction module comprises a frequency band filtering unit, a time-frequency transformation unit, and a feature extraction unit. The frequency band filtering unit is configured to receive the raw audio signal data output by the audio signal acquisition module, filter the signal in the frequency domain using a band-pass filter to filter out low-frequency mechanical vibration noise and high-frequency electromagnetic interference in the background environment, and retain leakage-related audio components within a predetermined frequency band. The time-frequency transformation unit is configured to perform short-time Fourier transform processing on the filtered audio signal to generate a time-frequency distribution map of the audio signal. The feature extraction unit is configured to extract the position of a sudden change point, the energy value of a high-frequency continuous segment, and the spectral slope feature from the time-frequency map, and output them in the form of a feature vector. The feature extraction unit comprises a spectral envelope analysis subunit and a feature output unit. The mutation point detection subunit is configured to extract high-energy mutation points along a time axis direction in a time-frequency spectrum, calculate a spectrum energy difference sequence of adjacent time frames, and determine an energy mutation position based on a set mutation threshold determine an energy mutation position; High frequency band identification subunit: for identifying high frequency band signals that continuously appear in the time-frequency spectrum, setting a frequency threshold , counting the average energy greater than , and screening continuous high frequency band segments that meet the condition of continuous time not less than the set length ; The spectral envelope analysis subunit is configured to extract the trend of spectral envelope changes in consecutive frames, construct an envelope change curve after normalizing the envelope energy, and calculate the change rate using the first derivative to output the spectral slope feature. The feature output unit is configured to concatenate the output sudden change point position, high-frequency continuous segment energy value, and spectral slope feature to form a fixed-dimensional feature vector. The leakage identification and positioning module comprises a leakage event identification unit, a signal strength analysis unit, and a leakage positioning calculation unit. The leakage event identification unit is configured to receive the feature vector output by the feature extraction unit, sequentially obtain the high-energy sudden change amplitude, the average energy value of the continuous high-frequency segment, and the spectral slope value from the feature vector, and compare them with the corresponding preset leakage identification threshold values. The signal strength analysis unit is configured to receive the filtered audio signals from each sensor in the audio signal acquisition module, calculate the signal strength values corresponding to each sensor, and record the spatial position coordinates of the corresponding sensors. The leakage positioning calculation unit is configured to, after receiving the leakage confirmation signal, derive the relative position coordinates of the leakage point based on the signal strength values and spatial positions of multiple adjacent sensors through a propagation attenuation model, and output the identification result of the leakage event and its corresponding spatial position information.

2. The audio based intelligent monitoring system for pipeline leak as claimed in claim 1 wherein, The audio signal acquisition module comprises a sensor arrangement unit, a signal sampling unit, and a data output unit. The sensor arrangement unit is configured to arrange a plurality of acoustic sensors along the pipeline at a preset interval, and to attach the sensors to the outer side of the pipeline wall and record the installation position coordinates of the sensors along the pipeline by using positioning tags. The signal sampling unit is configured to control the acoustic sensors to synchronously collect the vibration of the acoustic wave on the inner wall of the pipeline at a fixed sampling frequency, and to generate a continuous time-domain acoustic signal sequence. The data output unit is configured to mark the acoustic signal sequence collected by each sensor with a time stamp, to uniformly package the acoustic signal sequence in a format, and to output the original acoustic signal data.

3. The audio based intelligent monitoring system for pipeline leak as claimed in claim 1 wherein, The signal strength analysis unit includes: The envelope extraction subunit is configured to receive the filtered acoustic signal of each sensor in the acoustic signal acquisition module, to extract the instantaneous amplitude envelope of the acoustic signal by using Hilbert transform, and to generate a signal envelope sequence of each sensor channel. intensity calculating subunit: for calculating the whole section energy mean value of the signal envelope sequence of each sensor to obtain the acoustic frequency signal intensity value of the corresponding sensor ; The position mapping subunit is configured to call the preset sensor installation layout information, to obtain the three-dimensional spatial coordinate value corresponding to each sensor, and to bind and store the signal strength value calculated.

4. The audio based intelligent monitoring system for pipeline leak as claimed in claim 3 wherein, The leakage positioning calculation unit includes: The sensor pair screening subunit is configured to, after receiving the leakage confirmation signal, screen a plurality of adjacent sensors with signal strength values higher than a set threshold from all the sensors to form an effective sensing pair, and to retain the spatial position coordinates and the corresponding strength values. intensity difference calculation subunit: for calculating the difference between the audio signal intensity values of any pair of sensors and , and deducing the distance difference between the leakage point and the two sensors based on the energy attenuation relationship of the sound wave during propagation in the pipeline medium ; Position inversion derivation subunit: for deriving, based on multiple sets of corresponding spatial position coordinates, relative spatial position coordinates of the leakage points by least square method inversion ; The output result is that the spatial coordinates of the inversion obtained leakage point are taken as the positioning result of the leakage event, and the corresponding leakage confirmation information is output together to form the final leakage identification result and spatial position information.

5. The audio based intelligent monitoring system for pipeline leak as claimed in claim 1 wherein, The state evaluation and alarm module includes a leakage level determination unit, a state generation unit, and an alarm output unit; wherein: The leakage level determination unit is configured to calculate the leakage influence range and intensity level according to the leakage identification result and spatial position information output by the leakage identification and positioning module, in combination with the corresponding signal strength value and the multi-source sensor coverage, to determine the leakage event according to the preset evaluation standard, and to output the leakage level label. The state generation unit is configured to generate a standardized leakage state evaluation result according to the leakage level label and the leakage position coordinates, including the event number, the occurrence time, the leakage level, the spatial coordinates, and the current processing state label. The alarm output unit is configured to send the alarm information to the upper computer or the management terminal through the industrial communication interface after the leakage state evaluation result is generated.

6. The audio based intelligent monitoring system for pipeline leak as claimed in claim 5 wherein, The leakage level determination unit includes: The intensity index calculation subunit is configured to acquire the spatial position of the leakage point output by the leakage identification and positioning module, extract the signal intensity values of a plurality of sensors near the spatial position, and calculate the average intensity index of the leakage area ; Influence range estimation subunit: used to analyze signals whose strength values ​​exceed a preset strength threshold. The number of sensors and their spatial distribution range are used to calculate the radius of influence of the leak. ; The grade decision generation sub-unit is configured to match the calculated value with a preset leakage grade evaluation standard, determine the grade to which the leakage event belongs, and output a corresponding leakage grade label. With the calculated value, match in a preset leakage grade evaluation standard, determine the grade to which the leakage event belongs, and output a corresponding leakage grade label .

7. The method of intelligent monitoring of pipeline leak based on audio frequency, realized by the system of intelligent monitoring of pipeline leak based on audio frequency according to any one of claims 1-6, characterized in that, The method includes the following steps: S1: arranging a plurality of acoustic sensors at a preset position along the pipeline, collecting the original acoustic signal data in the pipeline operation process, and marking the spatial position information of each sensor; S2: performing band-pass filtering on the original acoustic signal obtained in S1 to suppress low-frequency mechanical interference and high-frequency electromagnetic noise, then calculating the high-energy mutation point, the continuous high-frequency band average energy value, and the frequency spectrum slope value to generate a feature vector; S3: comparing the feature vector generated in S2 with the preset leakage identification threshold in sequence, and if any feature dimension exceeds the corresponding threshold, outputting a leakage confirmation signal and determining the current detection frame as a leakage state. S4: After confirming the leakage state, extract the filtered signals of multiple sensors adjacent to the leakage point in space in S1, calculate the corresponding signal strength values, and call the spatial coordinate information as the subsequent positioning input; S5: Based on the signal strength values and spatial positions of the adjacent multiple sensors obtained in S4, construct a sound wave propagation attenuation model, deduce the distance difference from the leakage point to each sensor, and use the least squares method to inversely calculate the relative spatial position coordinates of the leakage point; S6: Receive the leakage confirmation signal and the leakage point coordinates output by S3 and S5, calculate the leakage level combined with the average leakage intensity and the influence radius, package the leakage state evaluation result, and output the alarm information to the upper computer or management terminal through the communication interface.

Images