A Method and System for Monitoring the Operation Status of Coal Chute Based on Distributed Acoustic Sensing

Abstract

This invention discloses a method and system for monitoring the operating status of a coal chuting pipeline based on distributed acoustic sensing. The method divides the coal chuting pipeline along its axial direction into several monitoring zones, and lays distributed acoustic sensing optical fibers along the outer wall of the pipeline along its axial direction. Continuous sampling is performed on the distributed acoustic sensing fibers along their length to obtain pre-processed acoustic signals within each time window of each monitoring zone. Frequency domain analysis is then performed, calculating the energy proportion of each frequency band and constructing feature vectors for each time window of each monitoring zone based on the main peak frequency, spectral entropy, spectral similarity, and impact amplitude characteristics in the spectral signal. An acoustic reference benchmark is established for each monitoring zone under normal coal flow impact conditions. The joint anomaly score for each time window of each monitoring zone is obtained by combining the feature vectors and the acoustic reference benchmark, and a joint temporal and spatial judgment is performed to determine the abnormal monitoring zone and anomaly type.

Description

Technical Field

[0001] This invention belongs to the field of distributed optical fiber sensing technology, specifically relating to a method and system for monitoring the operating status of coal chutes based on distributed acoustic sensing. Background Technology

[0002] Coal chutes are critical transfer structures in coal mines, thermal power plants, ports, and bulk material conveying systems. They are primarily used for gravity transport of bulk materials such as coal between different heights or conveying equipment. During actual operation, the coal stream falls rapidly along the inside of the chute under gravity, continuously impacting, rubbing, and colliding with the pipe walls, liners, bends, and outlet sections. This subjects the chutes to prolonged periods of high impact, high wear, and complex vibration. Any abnormal conditions such as localized blockages, arching, coal flow deviation, loose linings, or pipe wall wear will directly affect the continuous and stable operation of the material conveying system, and in severe cases, may cause equipment damage, downtime for maintenance, or even safety accidents.

[0003] In applications such as coal conveying systems in mines, thermal power plants, and port loading and unloading, coal chutes are typically characterized by dispersed locations, relatively enclosed structures, high dust levels, strong background noise, and frequent changes in operating conditions. Anomalies within coal chutes often occur in localized sections, and early-stage anomalies are not easily detected through external observation. Furthermore, different sections exhibit significant differences in vibration and acoustic response due to variations in installation angle, lining condition, coal flow impact location, and structural constraints. Therefore, the operating status of coal chutes has a crucial impact on the safety, stability, and operational efficiency of the entire bulk material conveying system.

[0004] Currently, the main methods for monitoring the operational status of coal chutes include manual inspection, video observation, temperature detection, pressure detection, and point-based vibration or acoustic sensing. While these methods can provide some status reference in practical applications, they are affected by factors such as on-site dust, noise, installation location, inspection cycle, and the pipeline's enclosed structure, making it difficult to comprehensively reflect the continuous operational status of different sections along the coal chutes. With the development of distributed fiber optic sensing technology, distributed acoustic sensing has gradually been applied to pipeline condition monitoring. By deploying sensing fibers along the pipeline, continuous acquisition and analysis of sound waves or vibration signals along the pipeline can be achieved.

[0005] However, the acoustic signal generation mechanisms of ordinary fluid pipelines and coal chutes differ significantly. In ordinary oil, gas, and water pipelines, acoustic signals primarily originate from the coupling between the continuous fluid medium and the pipe wall, while in coal chutes, acoustic signals are mainly formed by the impact, friction, collision, rolling, and deflection of discrete coal particles, as well as the response of the pipe wall structure. During normal operation, instantaneous impacts from large coal pieces, changes in coal flow rate, variations in the drop angle, and differences in local pipe wall structure can all cause significant fluctuations in the frequency, energy, and waveform of the acoustic signal; furthermore, abnormal conditions such as blockage, arching, deflection, and wear can lead to changes in the temporal, frequency, and spatial distribution of the acoustic response. Therefore, there is still a need for further improvement in the condition monitoring methods for coal chutes, which are characterized by strong impacts, high noise, and non-stationary operation. Summary of the Invention

[0006] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention proposes a method and system for monitoring the operating status of coal chutes based on distributed acoustic sensing.

[0007] To achieve the above objectives, according to one aspect of the present invention, a method for monitoring the operating status of a coal chuting pipeline based on distributed acoustic sensing is provided, comprising the following steps: The coal chute is divided into several measurement zones along the axial direction, and distributed acoustic sensing optical fibers are laid on the outer wall of the coal chute along the axial direction. The distributed acoustic sensing optical fibers are continuously sampled along the length direction to obtain the acoustic time-domain signal of each measurement zone. The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

[0008] According to the above method, the preprocessing steps for the acoustic time-domain signal of each measurement area after segmenting it according to a preset time window include detrending processing, filtering processing, normalization processing, windowing processing, and wavelet threshold denoising processing.

[0009] According to the above method, the step of dividing the spectrum into several frequency bands and calculating the energy proportion of each frequency band specifically includes: Based on the measured data, the effective frequency range was obtained. According to this effective frequency range, the spectrum was divided into a low-frequency structural response band, a mid-frequency friction and impact band, and a high-frequency particle collision band. The measured data are the preprocessed acoustic signals obtained in each time window of each measurement area during the period when each measurement area is in normal operation. The low-frequency structural response band is used to reflect changes in structural response caused by overall pipeline vibration, blockage, or arching; the medium-frequency friction and impact band is used to reflect the friction, rolling, or continuous impact state between the coal flow and the pipe wall or liner; and the high-frequency particle collision band is used to reflect the local collision of particulate coal flow, liner loosening, pipe wall wear, or abnormally strong impact state. Calculate the corresponding low-frequency, mid-frequency, and high-frequency energy proportions respectively.

[0010] Using the above method, the acoustic reference standards for each test area under normal coal flow impact conditions are established in the following way: For a specific test area, feature vectors from multiple time windows within the normal operating state are selected to form a normal feature sample set for that test area. Statistical analysis was performed on the normal characteristic sample set of the test area, and the reference center value and reference dispersion of each characteristic component were calculated. A robust statistical method was used to establish an acoustic reference benchmark under the normal coal flow impact state of the test area.

[0011] Using the above method, the reference center value is expressed as the mean, weighted average, or median, and the reference dispersion is expressed as the standard deviation, interquartile range, or absolute deviation of the median.

[0012] According to the above method, the fusion of the normalized deviations of each feature to obtain the joint anomaly score of each time window in each measurement area includes: the score is obtained by weighting and summing the normalized deviations of each feature, and the weight of the normalized deviation of each feature is determined by historical operating data and field calibration conditions.

[0013] According to the above method, the adaptive anomaly threshold of each test area is obtained based on the joint anomaly score distribution during the normal operation phase of the test area.

[0014] According to the above method, the joint temporal and spatial judgment of candidate abnormal test areas specifically includes: If the joint anomaly score of a candidate anomaly area exceeds the adaptive anomaly threshold of the candidate anomaly area a certain number of times within multiple consecutive time windows, it is determined that the candidate anomaly area has a continuous anomaly. Meanwhile, if the adjacent test areas of the candidate abnormal test area also have continuous anomalies within the same or a certain time window before and after, then the candidate abnormal test area is confirmed as an abnormal test area; otherwise, it is determined to be a transient impact or environmental interference.

[0015] According to the above method, the step of determining the anomaly type based on the corresponding feature vectors in the anomaly test area includes: If the proportion of low-frequency energy increases while the proportion of mid-frequency and high-frequency energy decreases, then the abnormality type is determined to be blockage or arching. If the proportion of high-frequency energy and the impact amplitude continue to increase, the abnormality type is determined to be pipe wall wear, liner loosening or local abnormal impact. If the energy distribution in multiple adjacent measurement areas is uneven over a long period of time, the anomaly type is determined to be either coal flow deviation or uneven material drop.

[0016] According to another aspect of the present invention, a coal chuting pipeline operation status monitoring system based on distributed acoustic sensing is provided, comprising: Light source generation module; Distributed acoustic sensing fiber optic cable is connected to the light source generation module via an optical circulator and laid on the outer wall of the coal chute along the axial direction of the coal chute; the coal chute is divided into several measurement zones along the axial direction. The data acquisition module, connected to the optical circulator, is used to perform photoelectric conversion and continuous sampling of the distributed acoustic sensing fiber along its length to obtain the acoustic time-domain signal of each measurement area. The signal processing module is used to complete the following steps: The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

[0017] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: By dividing the coal chutes into different monitoring zones and performing frequency domain analysis on the acoustic signals of each zone, the spectrum is divided into several frequency bands. The energy proportion of different frequency bands, along with the main peak frequency, spectral entropy, spectral similarity, and impact amplitude characteristics in the spectral signal, are used to construct feature vectors for each monitoring zone and time window. Acoustic reference benchmarks under normal coal flow impact conditions are established for different monitoring zones. Anomalies are determined by comparing real-time monitoring values ​​with the acoustic reference benchmarks. Furthermore, the feature vectors are strongly correlated with anomaly types, enabling stable identification and segment location of abnormal conditions such as coal chutes blockage, arching, flow deviation, and wear. This reduces false alarms caused by normal coal flow impact and environmental noise, improves the reliability of early anomaly identification, and enhances the targeted nature of operation and maintenance. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a distributed acoustic sensing system structure provided in an embodiment of the present invention.

[0019] Figure 2 This is a flowchart of a method provided in an embodiment of the present invention.

[0020] In the picture: 1-Laser generating module, 2-Optical circulator, 3-Distributed acoustic sensing fiber optic cable, 4-Coal chute, 5-Detector, 6-Signal processing module, 7-Driver, 8-Acquisition card. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0022] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0023] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified. In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0025] like Figure 1As shown, the distributed acoustic sensing system includes a laser generating module 1, an optical circulator 2, a distributed acoustic sensing fiber optic cable 3, a detector 5, a signal processing module 6, and a data acquisition card 8. The output of the laser generating module 1 is connected to the distributed acoustic sensing fiber optic cable 3 via the optical circulator 2. The distributed acoustic sensing fiber optic cable 3 is laid along the axial direction of the coal chute 4 on the outer wall of the chute 4. The returned acoustic time-domain signal is received by the detector 5 in the optical circulator 2 and converted into an electrical signal. This signal is then acquired by the signal processing module 6 via the data acquisition card 8, where it is processed to ultimately monitor the operating status of the coal chute 4. In some embodiments, the laser generating module 1 typically includes a laser, an acousto-optic modulator (AOM), and an erbium-doped fiber amplifier (EDFA) connected in sequence to output the laser light required by the distributed acoustic sensing fiber optic cable 3. In some embodiments, the acousto-optic modulator (AOM) and the data acquisition card 8 are driven by a driver 7. The acquisition of the acoustic time-domain signal is a conventional technique, and the specific parameter requirements of each part will not be repeated here.

[0026] The present invention aims to provide a method and system for monitoring the operating status of coal chutes based on distributed acoustic sensing, so as to realize online identification and measurement area determination of the operating status, abnormal conditions and potential faults of coal chutes.

[0027] To achieve the above objectives, according to one aspect of the present invention, this embodiment provides a method for monitoring the operating status of coal chutes based on distributed acoustic sensing, such as... Figure 2 As shown, it includes the following steps: S1. Sampling of acoustic time-domain signals and division of the test area: The coal chute is divided into several measurement zones along the axial direction, and distributed acoustic sensing optical fibers are laid on the outer wall of the coal chute along the axial direction. The distributed acoustic sensing optical fibers are continuously sampled along the length direction to obtain the acoustic time-domain signal of each measurement zone. S2. Segmentation and preprocessing of the acoustic wave time-domain signal: The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. S3. Feature extraction, constructing feature vectors: Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. S4. Establish acoustic reference standards under normal coal flow impact conditions in each test area: During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. S5. Combine anomaly scoring to obtain candidate anomaly detection areas: During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. S6. Determine the abnormal test area and abnormal type: The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

[0028] According to another aspect of the present invention, this embodiment also provides a coal chuting pipeline operation status monitoring system based on distributed acoustic sensing, including a light source generation module, a distributed acoustic sensing optical fiber, a data acquisition module, and a signal processing module.

[0029] The distributed acoustic sensing fiber is connected to the light source generation module via an optical circulator and laid along the axial direction of the coal chute on its outer wall; the coal chute is divided into several measurement zones along its axial direction. The data acquisition module is connected to the optical circulator and is used to perform photoelectric conversion and continuous sampling along the length of the distributed acoustic sensing fiber to obtain the acoustic time-domain signal of each measurement zone. Typically, the data acquisition module includes a detector and a data acquisition card.

[0030] The signal processing module is used to complete the following steps: The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

[0031] The invention will be further illustrated below with specific examples.

[0032] This embodiment provides a method for monitoring the operating status of coal chutes based on distributed acoustic sensing, including the following steps: S1. Sampling of acoustic time-domain signals and division of the test area: The coal chute is divided into several measurement zones along its axial direction; the distributed acoustic sensing fiber is continuously sampled along its length to obtain the acoustic time-domain signal of each measurement zone.

[0033] Distributed acoustic sensing optical fibers are laid on the outer wall of the coal chutes along the axial direction of the chutes, so that the distributed acoustic sensing optical fibers form a stable mechanical coupling relationship with the outer wall of the coal chutes, thereby enabling the sensing of acoustic signals generated by the coal chutes under the impact, friction, collision of coal flow and the response of the chutes structure.

[0034] The distributed acoustic sensing fiber is continuously sampled along its length using a distributed acoustic sensing system. Based on the spatial resolution of the distributed acoustic sensing system, the coal chuting pipeline is divided into several measurement zones. Let N be the total number of measurement zones along the coal chuting pipeline, and let the i-th measurement zone be denoted as [i-th zone]. Where i = 1, 2, ..., N. Each survey area They are distributed sequentially along the axial direction of the coal chute, and correspond one-to-one with different axial positions of the coal chute.

[0035] Based on the actual structural form of the coal chutes, the test area is... This mapping is further extended to structural regions such as the inlet section, straight pipe section, bend section, impact section, liner section, and outlet section, giving each test area not only the spatial location attributes of the optical fiber but also the structural attributes of the pipeline. Through this test area division and structural mapping, a spatial basis can be provided for subsequent acoustic feature modeling and anomaly segment localization in different test areas.

[0036] During the operation of the coal chutes, the coal flows down the inside of the chutes under the influence of gravity, and undergoes random impacts, friction, and collisions with the pipe walls, linings, and bending structures, causing the pipe structure to generate a time-varying acoustic response. These acoustic waves are transmitted through the pipe walls to the distributed acoustic sensing optical fibers, and corresponding time-domain acoustic signals are obtained in each measurement area. The time-domain acoustic signal collected in the i-th measurement area can be expressed as:

[0037] in, Indicates the i-th test area The collected acoustic wave time-domain signal, where t represents the sampling time and N represents the total number of survey areas.

[0038] Combination Figure 1 It can be seen that the distributed acoustic sensing fiber is laid along the axial direction of the coal chuting pipe, and each measurement area can form a corresponding spatial mapping relationship with the structure of the coal chuting pipe.

[0039] S2. Segmentation and preprocessing of the acoustic wave time-domain signal: The acoustic time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic signals of each measurement area within each time window.

[0040] In some embodiments, the acoustic time-domain signals of each measurement area obtained in S1 The signal is segmented according to a preset time window. The acoustic signal in the i-th measurement area within the j-th time window can be represented as... .

[0041] The segmented acoustic signal of the test area Preprocessing is performed to reduce the impact of environmental noise, low-frequency drift, differences in measurement area coupling, and occasional interference on subsequent feature extraction and anomaly identification. Specifically, this includes the following processing: First, detrending processing. The acoustic signals from each measurement area are detrended to remove low-frequency drift caused by the distributed acoustic sensing system, slow changes in pipeline structure, and baseline shifts caused by changes in environmental background.

[0042] Second, filtering. Based on the effective frequency range of the acoustic signal from the coal chutes, high-pass or band-pass filtering is applied to the acoustic signal in the measurement area to retain the effective frequency components related to coal flow impact, friction, collision, and structural response, while suppressing low-frequency environmental disturbances and high-frequency random noise.

[0043] Third, normalization processing. Amplitude or energy normalization is performed on the acoustic signals from different measurement areas to reduce amplitude scale differences caused by variations in fiber optic cable tightness, pipe wall thickness, liner condition, and coupling conditions.

[0044] Fourth, windowing processing. Windowing processing is applied to the segmented signal to reduce spectral leakage during frequency domain analysis and improve the stability of spectral characteristics.

[0045] Fifth, wavelet threshold denoising is performed to suppress non-coal flow impact noise and transient abnormal interference.

[0046] After the above processing, the preprocessed acoustic signal within the j-th time window of the i-th measurement area is obtained, which can be expressed as: This preprocessed signal is used for subsequent multi-band feature extraction.

[0047] S3. Feature extraction, constructing feature vectors: Frequency domain analysis was performed on the preprocessed acoustic signals in each time window of each measurement area to obtain the spectrum signal of the corresponding measurement area. The spectrum was divided into several frequency bands, and the energy proportion of each frequency band was calculated. Together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude features in the spectrum signal, the feature vector of each measurement area and time window was constructed.

[0048] In some embodiments, frequency domain analysis is performed on the preprocessed acoustic signals of each measurement area obtained in S2 to obtain the spectral representation of the corresponding measurement area. The spectrum of the i-th measurement area within the j-th time window is represented as follows: .

[0049] Considering that the acoustic signal of the coal chutes is mainly formed by the impact, friction, collision, rolling of coal flow and the response of the pipe wall structure, this invention extracts multi-frequency band impact characteristics that can reflect the operating status of the coal chutes.

[0050] Based on the measured data, the effective frequency range was obtained, and the spectrum was divided into a low-frequency structural response band, a mid-frequency friction and impact band, and a high-frequency particle collision band, denoted as follows:

[0051] The measured data consist of preprocessed acoustic signals obtained within each time window of each measurement area during the normal operation phase. Low-frequency structural response band. It mainly reflects the structural response changes caused by overall pipeline vibration, blockage, or arching; mid-frequency friction and impact band. It mainly reflects the friction, rolling, and continuous impact state between the coal flow and the pipe wall or lining; high-frequency particle collision frequency band. It mainly reflects the local collision of particulate coal flow, loosening of lining plates, wear of pipe walls, or abnormally strong impact.

[0052] Calculate the low-frequency energy in the j-th time window of the i-th measurement area respectively. Mid-frequency energy and high frequency energy :

[0053]

[0054]

[0055] Further calculation of the energy proportion of each frequency band , , :

[0056]

[0057]

[0058] By using energy proportion characteristics, the impact of amplitude differences in different measurement areas on feature extraction results can be reduced, and the changes in frequency band energy distribution under different operating conditions can be highlighted.

[0059] In addition to the multi-band energy ratio, the main peak frequency is also extracted. Spectral entropy Spectral similarity and impact amplitude characteristics Among them, the main peak frequency is used to characterize the frequency location where the main energy of the acoustic signal in the test area is concentrated; the spectral entropy is used to characterize the dispersion of the spectral energy distribution; the spectral similarity is used to characterize the degree of closeness between the current spectrum and the normal reference spectrum; the impact amplitude characteristics can be represented by the root mean square value, kurtosis, envelope kurtosis or peak factor, which are used to reflect the instantaneous and impulsive nature of coal flow impact.

[0060] Therefore, the multi-feature vector of the i-th test area and the j-th time window is constructed:

[0061] Through the above multi-band feature extraction process, the operating status of the coal chutes can be characterized from multiple perspectives, such as frequency band energy distribution, frequency position, spectral structure and impact pulse characteristics, providing a basis for subsequent acoustic benchmark modeling and anomaly judgment in the test area.

[0062] S4. Establish acoustic reference standards under normal coal flow impact conditions in each test area: During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired for multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area.

[0063] In some embodiments, during the period when the coal chutes are in normal operation, multiple feature vectors of multiple time windows in each measurement area are continuously acquired to establish an acoustic reference benchmark under normal coal flow impact conditions in each measurement area.

[0064] For the i-th test area, multiple time windows within the normal operation phase are selected to form a set of normal characteristic samples for that test area:

[0065] in, Let M represent the set of characteristic samples during the normal operation phase of the i-th test area, and M represent the number of time windows used to establish the normal acoustic reference.

[0066] Statistical analysis is performed on the set of normal feature samples to calculate the reference center value and reference dispersion for each feature component. The reference center value can be expressed as the mean, weighted average, or median, and the reference dispersion can be expressed as the standard deviation, interquartile range, or absolute deviation of the median.

[0067] To mitigate the impact of occasional large coal impacts or short-term operating condition fluctuations on the establishment of the normal reference benchmark, a robust statistical method is adopted. The normal reference center value of the q-th characteristic component in the i-th survey area is expressed as: The corresponding reference dispersion is expressed as Then the normal acoustic reference standard for the i-th test area It can be represented as:

[0068] Because the installation angle, structural constraints, lining condition, coal flow impact location, and pipe wall wear may differ in different test areas of the coal chutes, the normal acoustic response of each test area is not entirely the same. Therefore, this invention establishes a normal acoustic reference standard for each test area. Instead of using a unified baseline across the entire pipeline, this approach reduces the impact of structural differences in the survey area on anomaly detection results, improving the adaptability and reliability of anomaly identification across different survey areas.

[0069] S5. Combine anomaly scoring to obtain candidate anomaly detection areas: During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each time window in each measurement area are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each time window in each measurement area. The score is then compared with the adaptive anomaly threshold to obtain candidate anomaly measurement areas.

[0070] In some embodiments, during the real-time operation of the coal chutes, multiple feature vectors of the i-th measurement area and the j-th time window are acquired. and compared with the normal acoustic reference standard of the test area. Compare and calculate the deviation of each feature from the normal state.

[0071] For the q-th feature component, its normalized deviation It can be represented as:

[0072] in, Indicates the current eigenvalue. This represents the reference center value of the feature under normal conditions. Indicates the reference dispersion. To prevent constants with a denominator of zero.

[0073] The deviations of each feature are weighted and fused to obtain the joint anomaly score for the i-th measurement area and the j-th time window. :

[0074] in, Let represent the weight of the q-th feature, and satisfy:

[0075] The weighting principle is as follows: features with obvious abnormal responses and small fluctuations under normal conditions are given larger weights, features that are susceptible to environmental noise are given smaller weights, and different weight combinations can be used for different types of anomalies. The weights are determined by historical operating data and on-site calibration conditions.

[0076] Based on the joint anomaly score distribution of each monitoring area during normal operation, an adaptive anomaly threshold is established for the corresponding monitoring area:

[0077] in, This represents the set of joint anomaly scores for the i-th survey area during its normal operation phase. Indicates the preset quantile.

[0078] When the joint anomaly score of the current test area satisfies:

[0079] The area in question is then marked as a candidate anomalous area within the specified time window and proceeds to subsequent temporal and spatial continuity checks.

[0080] S6. Determine the abnormal test area and abnormal type: The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

[0081] In some embodiments, the candidate abnormal test areas obtained in S5 are subjected to joint temporal and spatial judgment. If the joint abnormality score of a certain test area exceeds the adaptive abnormality threshold of that test area a set requirement multiple times within consecutive time windows, then the test area is considered to have a persistent abnormality. The judgment condition can be expressed as follows:

[0082] Where I(·) is the indicator function, This represents the minimum number of times the abnormal condition is met.

[0083] Spatial judgment is further made by combining the abnormal score changes of adjacent measurement areas. If a candidate abnormal measurement area and its adjacent measurement areas show continuous abnormal responses within a similar time period, the measurement area is confirmed as an abnormal measurement area; if the abnormality only appears in a single time window or an isolated measurement area, it is determined to be a transient impact or environmental interference, and no fault alarm is output.

[0084] After identifying the abnormal measurement area, the anomaly type is determined based on changes in multi-band energy, main peak frequency shift, spectral entropy, spectral similarity, and impact amplitude. If low-frequency energy increases while mid-to-high-frequency energy decreases, it can be identified as blockage or arching; if the proportion of high-frequency energy and impact amplitude continue to increase, it can be identified as pipe wall wear, liner loosening, or localized abnormal impact; if the energy distribution in multiple adjacent measurement areas is uneven for a long period, it can be identified as coal flow deviation or uneven material drop.

[0085] When multiple adjacent survey areas are confirmed as anomalous, they are merged into a continuous anomalous section. The set of continuous anomalous survey areas can be represented as:

[0086] in, Represents a set of continuous anomaly survey areas. and These represent the starting and ending survey areas of the abnormal section, respectively.

[0087] The final output includes the time of the anomaly, the anomaly area number, the range of the anomaly segment, the anomaly type, and the joint anomaly score, providing a basis for operations and maintenance personnel to conduct targeted inspections.

[0088] This invention, based on distributed acoustic sensing technology, lays sensing optical fibers on the outer wall of the coal chuting pipeline and divides the pipeline into multiple measurement zones. This enables continuous acquisition and online analysis of acoustic signals from different sections of the coal chuting pipeline, allowing for timely detection of local anomalies and overcoming the limitations of manual inspections and point-based sensing, which suffer from limited coverage and delayed response. By extracting features such as multi-band energy, peak frequency, spectral entropy, spectral similarity, and impact amplitude to construct a feature vector, and establishing normal acoustic reference benchmarks for each measurement zone, this invention can more fully characterize the impact, friction, collision, and structural response changes of coal flow in the coal chuting pipeline, reducing misjudgments caused by single features. Furthermore, this invention combines joint anomaly scoring, temporal persistence, and spatial continuity to confirm candidate anomaly measurement zones, reducing false alarms caused by instantaneous large coal impacts, occasional noise, or environmental disturbances. Simultaneously, it can merge adjacent anomaly measurement zones into continuous fault sections, outputting the anomaly measurement zone, anomaly type, and fault section range, providing a basis for maintenance personnel to conduct targeted inspections.

[0089] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0090] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0091] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the operating status of coal chutes based on distributed acoustic sensing, characterized in that: Includes the following steps: The coal chute is divided into several measurement zones along the axial direction, and distributed acoustic sensing optical fibers are laid on the outer wall of the coal chute along the axial direction. The distributed acoustic sensing optical fibers are continuously sampled along the length direction to obtain the acoustic time-domain signal of each measurement zone. The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

2. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The preprocessing steps for the acoustic time-domain signals of each measurement area, after segmenting them according to a preset time window, include detrending processing, filtering processing, normalization processing, windowing processing, and wavelet threshold denoising processing.

3. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The step of dividing the spectrum into several frequency bands and calculating the energy proportion of each frequency band includes: Based on the measured data, the effective frequency range was obtained. According to this effective frequency range, the spectrum was divided into a low-frequency structural response band, a mid-frequency friction and impact band, and a high-frequency particle collision band. The measured data are the preprocessed acoustic signals obtained in each time window of each measurement area during the period when each measurement area is in normal operation. The low-frequency structural response band is used to reflect changes in structural response caused by overall pipeline vibration, blockage, or arching; the medium-frequency friction and impact band is used to reflect the friction, rolling, or continuous impact state between the coal flow and the pipe wall or liner; and the high-frequency particle collision band is used to reflect the local collision of particulate coal flow, liner loosening, pipe wall wear, or abnormally strong impact state. Calculate the corresponding low-frequency, mid-frequency, and high-frequency energy proportions respectively.

4. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The acoustic reference standards for each test area under normal coal flow impact conditions were established using the following methods: For a specific test area, feature vectors from multiple time windows within the normal operating state are selected to form a normal feature sample set for that test area. Statistical analysis was performed on the normal characteristic sample set of the test area, and the reference center value and reference dispersion of each characteristic component were calculated. A robust statistical method was used to establish an acoustic reference benchmark under the normal coal flow impact state of the test area.

5. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 4, characterized in that: The reference center value is expressed as the mean, weighted average, or median, and the reference dispersion is expressed as the standard deviation, interquartile range, or absolute deviation of the median.

6. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The process of fusing the normalized deviations of each feature to obtain the joint anomaly score for each time window in each measurement area includes: weighting and summing the normalized deviations of each feature, with the weights of the normalized deviations of each feature determined by historical operating data and on-site calibration conditions.

7. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The adaptive anomaly threshold for each measurement area is obtained based on the joint anomaly score distribution during the normal operation phase of that measurement area.

8. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 1, characterized in that: The joint temporal and spatial judgment of candidate abnormal test areas specifically includes: If the joint anomaly score of a candidate anomaly area exceeds the adaptive anomaly threshold of the candidate anomaly area a certain number of times within multiple consecutive time windows, it is determined that the candidate anomaly area has a continuous anomaly. Meanwhile, if the adjacent test areas of the candidate abnormal test area also have continuous anomalies within the same or a certain time window before and after, then the candidate abnormal test area is confirmed as an abnormal test area; otherwise, it is determined to be a transient impact or environmental interference.

9. The method for monitoring the operating status of coal chutes based on distributed acoustic sensing according to claim 3, characterized in that: The step of determining the anomaly type based on the corresponding feature vector in the anomaly test area includes: If the proportion of low-frequency energy increases while the proportion of mid-frequency and high-frequency energy decreases, then the abnormality type is determined to be blockage or arching. If the proportion of high-frequency energy and the impact amplitude continue to increase, the abnormality type is determined to be pipe wall wear, liner loosening or local abnormal impact. If the energy distribution in multiple adjacent measurement areas is uneven over a long period of time, the anomaly type is determined to be either coal flow deviation or uneven material drop.

10. A coal chuting pipeline operation status monitoring system based on distributed acoustic sensing, characterized in that: include: Light source generation module; Distributed acoustic sensing fiber optic cable is connected to the light source generation module via an optical circulator and laid on the outer wall of the coal chute along the axial direction of the coal chute; the coal chute is divided into several measurement zones along the axial direction. The data acquisition module, connected to the optical circulator, is used to perform photoelectric conversion and continuous sampling of the distributed acoustic sensing fiber along its length to obtain the acoustic time-domain signal of each measurement area. The signal processing module is used to complete the following steps: The acoustic wave time-domain signals of each measurement area are segmented according to a preset time window and then preprocessed to obtain the preprocessed acoustic wave signals of each measurement area within each time window. Frequency domain analysis is performed on the preprocessed acoustic signal to obtain the spectrum signal of the corresponding test area. The spectrum signal is divided into several frequency bands, and the energy ratio of each frequency band is calculated. The energy ratio of each frequency band, together with the main peak frequency, spectral entropy, spectral similarity and impact amplitude characteristics in the spectrum signal, are used to construct the feature vector of each time window of each test area. During the period when each test area is in normal operation, multiple feature vectors of each test area are continuously acquired in multiple time windows to establish an acoustic reference benchmark under normal coal flow impact conditions in each test area. During the real-time operation of the coal chutes, the actual values ​​of the feature vectors of each measurement area and time window are obtained and compared with the acoustic reference benchmark. The normalized deviation of each feature in the feature vector is calculated. The normalized deviation of each feature is fused to obtain the joint anomaly score of each measurement area and time window. The score is then compared with the adaptive anomaly threshold to obtain the candidate anomaly measurement areas. The candidate abnormal test areas are jointly judged in time and space to identify the abnormal test areas, and the abnormality type is determined based on the corresponding feature vectors in the abnormal test areas.

Images