High-voltage electrical equipment inspection method and system based on microphone array measurement

Abstract

The invention relates to the technical field of power equipment inspection, in particular to a high-voltage electrical equipment inspection method and system based on microphone array measurement. The method comprises the following steps: acquiring array sound signal data, analyzing the array sound signal data, and determining sound signal feature information; obtaining a layered structure parameter set and an insulating medium parameter set of the target electrical equipment, performing targeted compensation and correction on the sound signal feature information according to the layered structure parameter set and the insulating medium parameter set, and determining abnormal voiceprint feature information; and generating abnormal component information according to the abnormal voiceprint feature information, and determining and outputting an abnormal detection report. According to the invention, non-contact real-time detection of the high-voltage electrical equipment in operation is realized, the adaptability of the detection process to the high-voltage electrical equipment which is internally provided with a multi-layer composite structure and is filled with an insulating medium is enhanced, and the positioning accuracy of a fault position in the equipment is improved.

Description

Technical Field

[0001] This application relates to the technical field of power equipment inspection, and in particular to a high-voltage electrical equipment inspection method and system based on microphone array measurement. Background Art

[0002] Acoustic imaging technology is a technology that uses the propagation characteristics of sound waves and combines signal processing technology to generate images of the sound field distribution inside or on the surface of an object. By capturing the reflection, scattering, transmission, or radiation characteristics of sound waves, acoustic information is converted into a visual image, which is widely used in the industrial inspection process.

[0003] However, when the existing acoustic imaging technology is used to detect abnormalities in high-voltage electrical equipment with multi-layer complex structures and dielectric isolation characteristics, it is difficult to accurately locate the internal fault positions of the equipment, resulting in the difficulty of guaranteeing the accuracy and reference value of the detection results. Summary of the Invention

[0004] This application provides a high-voltage electrical equipment inspection method and system based on microphone array measurement to solve the above technical problems.

[0005] In a first aspect, this application provides a high-voltage electrical equipment inspection method based on microphone array measurement, and the method includes: Obtain the array acoustic signal data of the target electrical equipment, analyze the array acoustic signal data, and determine the acoustic signal characteristic information; The acoustic signal characteristic information is a set of characteristics extracted from the array acoustic signal data and capable of characterizing the operating state of the target equipment; Obtain the layered structure parameter set and the insulating medium parameter set of the target electrical equipment, and perform targeted compensation and correction on the acoustic signal characteristic information according to the layered structure parameter set and the insulating medium parameter set to determine the abnormal acoustic pattern characteristic information; The abnormal acoustic pattern characteristic information is the sound characteristic information used to map the local abnormal discharge situation inside the target electrical equipment; Generate abnormal component information according to the abnormal acoustic pattern characteristic information, and determine and output an abnormal detection report.

[0006] Through this solution, based on acoustic imaging technology, according to the characteristic that abnormal sound signals will be generated during partial discharge of high-voltage electrical equipment, by analyzing the array acoustic signal data, the acoustic signal feature information used to characterize the current operating state of the high-voltage electrical equipment is extracted from the array acoustic signal data, and this is used as the data basis for judging whether there is partial discharge abnormality in the high-voltage electrical equipment, so as to realize non-contact real-time detection of the operating high-voltage electrical equipment. On this basis, the influence of the internal layered structure and insulating medium of the target electrical equipment on the sound signal is used as the correction basis for the acoustic signal feature information, and the acoustic signal feature information is compensated and corrected specifically to obtain the abnormal sound pattern feature information that can accurately characterize the actual partial discharge situation inside the electrical equipment, enhancing the adaptability of the detection process to high-voltage electrical equipment with a multi-layer composite structure and filled with insulating medium inside, improving the positioning accuracy of the internal fault location of the equipment, generating abnormal component information according to the abnormal sound pattern feature information, and providing the corresponding visual abnormal detection report to the corresponding maintenance personnel to improve the reference value of the detection result.

[0007] Optionally, the array acoustic signal data is obtained through a circular microphone array; The number of microphones in the circular microphone array is not less than 128; The acoustic measurement frequency range of the circular microphone array is 2 kHz - 40 kHz.

[0008] Through this solution, the array mode of the microphones, the number of microphones, and the acoustic measurement frequency range are restricted to meet the sound source localization requirements under the complex three-dimensional structure of high-voltage electrical equipment, improve the sound source localization accuracy, suppress environmental noise, increase the signal-to-noise ratio, and thus improve the comprehensiveness and accuracy of the detection process.

[0009] Optionally, analyzing the array acoustic signal data to determine the acoustic signal feature information includes: Extracting the continuous frame features of the array acoustic signal data according to the preset unit sound frame time; According to the continuous frame features, performing frame division processing on the array acoustic signal data, and extracting the acoustic signal time-domain features, acoustic signal frequency-domain features, and time-frequency joint non-stationary features; Based on the acoustic signal time-domain features, the acoustic signal frequency-domain features, and the time-frequency joint non-stationary features, separating the mechanical vibration acoustic signals and partial discharge acoustic signals in the array acoustic signal data according to the mechanical vibration acoustic features and partial discharge acoustic features; Constructing the acoustic signal feature information according to the partial discharge acoustic signals; The mechanical vibration acoustic features are mainly characterized by low-frequency harmonic groups and show periodic attenuation characteristics; The partial discharge acoustic features are characterized by high-frequency pulse clusters and have random burst characteristics.

[0010] Through this scheme, the array acoustic signal data is framed and processed, and the time domain characteristics, frequency domain characteristics and time-frequency joint non-stationary characteristics of the acoustic signal are extracted respectively. On this basis, according to the mechanical vibration acoustic characteristics and the partial discharge acoustic characteristics, the separation between the mechanical vibration acoustic signal and the partial discharge acoustic signal in the array acoustic signal data is realized, so as to construct the acoustic signal feature information according to the partial discharge acoustic signal, so that the acoustic signal feature information points to the abnormal discharge situation inside the equipment, avoids the confusion between the vibration caused by the operation of the equipment itself and the partial discharge anomaly, and prevents invalid maintenance or delayed repair due to misjudgment.

[0011] Optionally, the layered structure parameter set includes structure layer thickness distribution information, structure material information and geometric topology information; The insulating medium parameter set includes medium type, medium dynamic viscosity coefficient, medium sound velocity propagation characteristics and medium pressure reference.

[0012] Through this scheme, a layered structure parameter set is constructed according to the structural layer thickness distribution information, structural material information and geometric topology information. The influence of the multi-layer structure inside the high-voltage electrical equipment on the partial discharge acoustic signal is reflected from the three dimensions of material properties, material thickness and geometric shape. At the same time, an insulating medium parameter set is constructed according to the medium type, medium dynamic viscosity coefficient, medium sound speed propagation characteristics and medium pressure reference, providing a reliable data basis for a comprehensive analysis of the influence of insulating media on acoustic signals.

[0013] Optionally, performing targeted compensation and correction on the acoustic signal characteristic information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal voiceprint characteristic information includes: According to the layered structure parameter set, a physical field simulation model based on multi-layer geometric structure sound propagation is constructed; Determining the reflection-transmission attenuation coefficient of the partial discharge acoustic signal in different layered structures according to the physical field simulation model; According to the insulating medium parameter set, performing phase compensation correction on the propagation path of the partial discharge acoustic signal to determine the dielectric absorption coefficient; According to the reflection-transmission attenuation coefficient and the medium absorption coefficient, targeted compensation and correction are performed on the partial discharge acoustic signal to determine the abnormal soundprint feature information.

[0014] Through this solution, a physical field simulation model is constructed according to the layered structure parameter set to simulate the negative impact on the propagation process of acoustic signals between different layered structures inside high-voltage electrical equipment, and the corresponding reflection-transmission attenuation coefficient is obtained by fitting. According to the insulating medium parameter set, the propagation characteristics of acoustic signals in the insulating medium are analyzed, and the medium absorption coefficient reflecting the influence of the insulating medium on the propagation of acoustic signals is obtained by fitting. The attenuation characteristics of acoustic signals under the dual influence are characterized by the reflection-transmission attenuation coefficient and the medium absorption coefficient, and the actual sound signal characteristics caused by partial discharge are deduced in reverse, accurately locating the abnormal sound pattern characteristic information, and improving the judgment accuracy of local abnormal discharge.

[0015] Optionally, constructing a physical field simulation model based on the acoustic propagation of a multi-layer geometric structure according to the layered structure parameter set includes: According to the structural layer thickness distribution information and the geometric topology information, a three-dimensional space grid model is constructed, the geometric characteristics of each layer of structure are parametrically characterized, and several hierarchical grid parameter nodes are determined; Based on the structural material information, the acoustic impedance parameters of the corresponding structural material are loaded at each hierarchical grid parameter node, and according to the acoustic impedance parameters, the time delay characteristics, attenuation characteristics and superposition characteristics of the partial discharge acoustic signal in each propagation path are simulated and marked by the ray tracing algorithm, and the physical field simulation model is constructed.

[0016] Through this solution, a three-dimensional space grid model containing hierarchical grid parameter nodes is constructed according to the structural layer thickness distribution information and the geometric topology information, realizing the sound field simulation of the three-dimensional irregular structure inside the high-voltage electrical equipment, eliminating the source location offset caused by the two-dimensional model. On this basis, based on the acoustic impedance parameters reflected in the structural material information, the time delay characteristics, attenuation characteristics and superposition characteristics of the partial discharge acoustic signal in each propagation path are simulated and marked by the ray tracing algorithm, and the physical field simulation model is constructed, realizing the high-precision simulation of the acoustic wave propagation characteristics under the complex structure of the high-voltage equipment, providing a reliable numerical basis for the physical field compensation of abnormal sound pattern characteristics, and significantly improving the location accuracy and diagnostic reliability of partial discharge detection.

[0017] Optionally, compensating and correcting the phase of the propagation path of the partial discharge acoustic signal according to the insulating medium parameter set to determine the medium absorption coefficient includes: According to the medium type and the medium sound speed propagation characteristics, a basic dynamic propagation speed field model of the partial discharge acoustic signal in the insulating medium is constructed; Based on the basic dynamic propagation speed field model, the viscous phase delay of the partial discharge acoustic signal in the insulating medium is simulated according to the medium dynamic viscosity coefficient and the medium pressure reference; The viscous phase delay is positively correlated with the dynamic viscosity coefficient of the medium; The viscous phase delay is negatively correlated with the medium pressure reference; The viscous phase delay is normalized to determine the medium absorption coefficient.

[0018] Through this scheme, a basic dynamic propagation velocity field model is constructed according to the medium type and the sound speed propagation characteristics of the medium to reflect the basic propagation characteristics of the sound signal in the current insulating medium. On this basis, the dynamic viscosity coefficient of the medium and the medium pressure reference are introduced to simulate the viscous phase delay of the local discharge sound signal under the dual influence of the viscosity and pressure of the insulating medium, and the viscous phase delay is normalized to obtain the medium absorption coefficient to accurately reflect the actual propagation state of the sound signal in the insulating medium, thereby improving the accuracy of subsequent positioning analysis of the local discharge position.

[0019] Optionally, generating abnormal component information according to the abnormal voiceprint feature information includes: After determining several levels of grid parameter nodes in the physical field simulation model, the reflection-transmission attenuation coefficient and the medium absorption coefficient are substituted into an inverse ray tracing algorithm to invert the sound source propagation path, determine the coordinate point of the sound source of the partial discharge sound signal in the three-dimensional space grid model, and generate the three-dimensional coordinates of the abnormal position; Based on the geometric topology information, determining the relative position of the abnormal equipment component and the partial discharge according to the three-dimensional coordinates of the abnormal position; According to the relative position of the abnormal equipment component and the partial discharge, a preset equipment abnormality information database is retrieved to determine abnormality troubleshooting suggestions; The abnormal component information is constructed based on the abnormal equipment component, the relative position of the partial discharge and the abnormality troubleshooting suggestion.

[0020] Through this solution, based on the physical field simulation model, according to the reflection-transmission attenuation coefficient and the dielectric absorption coefficient, the inverse ray tracing algorithm is used to deduce the propagation path of the currently collected partial discharge acoustic signal, and then locate the sound source of the partial discharge acoustic signal inside the high-voltage electrical equipment, and obtain the three-dimensional coordinates of the abnormal position. Combined with the geometric topological information, the relative position of the abnormal equipment components and the partial discharge is determined, so as to retrieve the preset equipment abnormality information library, determine the abnormality troubleshooting suggestions, integrate the abnormal equipment components, the relative position of the partial discharge and the abnormality troubleshooting suggestions, and construct the abnormal component information, so as to achieve accurate positioning of the abnormal discharge position inside the high-voltage electrical equipment, and provide corresponding abnormality troubleshooting suggestions, effectively improving the inspection and maintenance efficiency and effectiveness of subsequent maintenance personnel.

[0021] Optionally, the method further includes: Obtain the maintenance feedback information of the maintenance personnel, analyze the maintenance feedback information of the maintenance personnel, and determine whether the maintenance personnel adopt the abnormal troubleshooting suggestions; If the maintenance personnel adopt the corresponding abnormal troubleshooting suggestions, increase the influence factor of the abnormal troubleshooting suggestions; If the maintenance personnel do not adopt the corresponding abnormal troubleshooting suggestions, incorporate the troubleshooting process in the maintenance feedback information of the maintenance personnel into the troubleshooting suggestion list corresponding to the relative position of the partial discharge under the corresponding abnormal equipment component in the preset equipment abnormal information library, and set its corresponding influence factor to the preset influence reference value; The abnormal troubleshooting suggestions in the troubleshooting suggestion list are sorted according to the numerical values of their corresponding influence factors.

[0022] Through this solution, based on the collected maintenance feedback information of the maintenance personnel, using the adjustment mechanism of the influence factor, dynamically update the abnormal troubleshooting suggestions in the abnormal troubleshooting suggestion list and adjust the recommendation priorities, establish a complete closed loop of "suggestion generation - on-site verification - knowledge evolution", enable the abnormal troubleshooting system to have the ability of continuous self-optimization, adapt to complex actual situations, and further improve the maintenance efficiency and maintenance effect of the maintenance personnel.

[0023] In a second aspect, the present application provides a high-voltage electrical equipment inspection system based on microphone array measurement, and the system includes: A feature analysis module, configured to obtain array acoustic signal data, analyze the array acoustic signal data, and determine acoustic signal feature information; A feature correction module, configured to obtain a layered structure parameter set and an insulating medium parameter set of a target electrical equipment, and perform targeted compensation and correction on the acoustic signal feature information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal acoustic fingerprint feature information; An abnormal positioning module, configured to generate abnormal component information according to the abnormal acoustic fingerprint feature information, and determine and output an abnormal detection report. Brief Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1 A schematic diagram of an application scenario provided by an embodiment of the present application; Figure 2Flowchart of a high-voltage electrical equipment inspection method based on microphone array measurement provided by an embodiment of the present application; Figure 3 Structural schematic diagram of a high-voltage electrical equipment inspection system based on microphone array measurement provided by an embodiment of the present application. Detailed implementation manners

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without making creative efforts shall fall within the protection scope of the present application.

[0027] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after, unless otherwise specified.

[0028] The following further describes the embodiments of the present application in detail with reference to the accompanying drawings of the specification.

[0029] However, when the existing acoustic imaging technology is used to detect abnormalities in high-voltage electrical equipment with multi-layer complex structures and dielectric isolation characteristics, it is difficult to accurately locate the internal fault positions of the equipment, resulting in difficulties in ensuring the accuracy and reference value of the detection results.

[0030] Based on this, the present application provides a high-voltage electrical equipment inspection method and system based on microphone array measurement. Based on acoustic imaging technology, according to the characteristics of abnormal sound signals generated when high-voltage electrical equipment is partially discharged, by analyzing the array sound signal data, the sound signal feature information used to characterize the current operating state of the high-voltage electrical equipment is extracted from the array sound signal data, which is used as the data basis for judging whether the high-voltage electrical equipment has partial discharge anomalies, and non-contact real-time detection of the high-voltage electrical equipment in operation is realized. On this basis, the influence of the internal layered structure and insulating medium of the target electrical equipment on the sound signal is used as the correction basis for the sound signal feature information, and the sound signal feature information is compensated and corrected in a targeted manner to obtain abnormal soundprint feature information that can accurately characterize the actual partial discharge situation inside the electrical equipment, so that the detection process is more adaptable to high-voltage electrical equipment with a multi-layer composite structure and filled with insulating medium, and the positioning accuracy of the internal fault position of the equipment is improved. According to the abnormal soundprint feature information, abnormal component information is generated, and the corresponding visual abnormal detection report is provided to the corresponding maintenance personnel, so as to improve the reference value of the detection results.

[0031] Figure 1 A schematic diagram of an application scenario provided by the present application. In the process of detecting abnormal partial discharge of high-voltage electrical equipment, the method provided by the present application is applied to improve the accuracy of abnormal partial discharge detection of high-voltage electrical equipment with multi-layer complex structures and dielectric isolation characteristics.

[0032] Specifically, the method of the present application is applied to any server, which communicates with a microphone array, and obtains array acoustic signal data provided by the microphone array through the server. Based on acoustic imaging technology, according to the characteristics of abnormal sound signals generated when high-voltage electrical equipment is partially discharged, the array acoustic signal data is analyzed to extract acoustic signal feature information used to characterize the current operating state of the high-voltage electrical equipment, and this is used as the data basis for judging whether there is a partial discharge abnormality in the high-voltage electrical equipment, so as to realize non-contact real-time detection of the high-voltage electrical equipment in operation. On this basis, the influence of the internal layered structure and insulating medium of the target electrical equipment on the sound signal is used as the basis for correcting the acoustic signal feature information, and the acoustic signal feature information is targetedly compensated and corrected to obtain abnormal soundprint feature information that can accurately characterize the actual partial discharge situation inside the electrical equipment, so that the detection process is more adaptable to high-voltage electrical equipment with a multi-layer composite structure and filled with insulating medium, and the positioning accuracy of the internal fault position of the equipment is improved. According to the abnormal soundprint feature information, abnormal component information is generated, and the corresponding visual abnormality detection report is provided to the corresponding maintenance personnel, so as to improve the reference value of the detection result.

[0033] For specific implementation methods, please refer to the following embodiments.

[0034] Figure 2 This is a flowchart of a high-voltage electrical equipment inspection method based on microphone array measurement provided in an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201: Acquire array acoustic signal data, analyze the array acoustic signal data, and determine acoustic signal feature information.

[0035] The array acoustic signal data may be a set of acoustic signal data generated during the operation of the target high-voltage electrical equipment and collected by a microphone array.

[0036] The acoustic signal feature information may be a feature set extracted from the array acoustic signal data and can be used to characterize the operating status of the device.

[0037] Specifically, high-voltage electrical equipment is a key equipment node in the process of power transmission. The detection and maintenance of periodic partial discharge anomalies of high-voltage electrical equipment are the basis for maintaining a continuous and good supply of electricity. However, due to the presence of high-voltage currents during the operation of high-voltage electrical equipment, the detection of high-voltage electrical equipment in operation is highly dangerous, and power-off detection will directly affect the continuity of power supply. Based on acoustic imaging technology, according to the characteristics of abnormal sound signals generated when high-voltage electrical equipment locally discharges, the sound signals of high-voltage electrical equipment in operation are collected through a microphone array to obtain array sound signal data. By analyzing the array sound signal data, the sound signal feature information in the array sound signal data used to characterize the current operating status of the high-voltage electrical equipment is extracted, which is used as the data basis for further judging whether there is a partial discharge anomaly in the high-voltage electrical equipment, thereby realizing non-contact real-time detection of high-voltage electrical equipment in operation.

[0038] S202, obtaining a layered structure parameter set and an insulating medium parameter set of the target electrical equipment, and performing targeted compensation and correction on the acoustic signal characteristic information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal voiceprint characteristic information.

[0039] The target electrical device may be a high voltage electrical device currently being inspected.

[0040] The layered structure parameter set may be a parameter set used to characterize the internal layered device structure characteristics of the high-voltage electrical device, and the layered structure parameter set may be provided by the device designer.

[0041] The insulating medium parameter set may be a set of parameters used to characterize the characteristics of the insulating medium filled inside the high-voltage electrical equipment, and the insulating medium parameter set may be provided by the equipment designer.

[0042] The abnormal voiceprint feature information may be sound feature information used to map the local abnormal discharge situation inside the high-voltage electrical equipment.

[0043] Specifically, when the existing acoustic imaging technology performs local abnormal discharge detection on high-voltage electrical equipment with a multi-layer composite structure and filled with insulating medium (such as transformers, enclosed gas insulated switches, etc.), the collected sound signals are affected by the attenuation caused by the multi-layer composite structure inside the equipment and the distortion caused by the insulating medium, resulting in large fluctuations in the accuracy of the detection results. It is difficult to accurately locate the specific abnormal discharge position inside the equipment, and it is impossible to provide equipment maintenance personnel with detection results with high reference value. This scheme collects the internal structure parameter set and the insulating medium parameter set of the target electrical equipment, and uses the influence of the internal layered structure and insulating medium of the target electrical equipment on the sound signal as the basis for correcting the characteristic information of the sound signal. Targeted compensation and correction are performed on the characteristic information of the sound signal to obtain abnormal soundprint characteristic information that can accurately characterize the actual local discharge situation inside the electrical equipment.

[0044] S203. Generate abnormal component information based on abnormal voiceprint feature information, and determine and output an abnormal detection report.

[0045] The abnormal component information may include the location of the faulty equipment component, the relative location of the partial discharge, and abnormality troubleshooting suggestions, which are of high reference value to equipment maintenance personnel.

[0046] The anomaly detection report may be a visual report document corresponding to the anomaly component information.

[0047] Specifically, the traditional detection method only outputs a simple alarm signal, lacks the precise positioning of the faulty components and maintenance guidance, resulting in low maintenance efficiency. This step uses reverse sound source tracking and equipment topology mapping technology to map and analyze the abnormal soundprint feature information, and generates abnormal component information including the location of the faulty equipment component, the relative position of the partial discharge, and abnormal troubleshooting suggestions, which is of high reference value to equipment maintenance personnel. The abnormal component information is visualized through data visualization technology to generate a corresponding abnormal detection report. The human-computer interaction equipment components, such as high-definition display screens, are used to provide the abnormal detection report to the corresponding equipment maintenance personnel for reference.

[0048] Through this solution, based on acoustic imaging technology, according to the characteristic that abnormal sound signals will be generated during partial discharge of high-voltage electrical equipment, by analyzing the array acoustic signal data, the acoustic signal characteristic information used to characterize the current operating state of the high-voltage electrical equipment is extracted from the array acoustic signal data, and this is used as the data basis for judging whether there is partial discharge abnormality in the high-voltage electrical equipment, so as to realize non-contact real-time detection of the operating high-voltage electrical equipment. On this basis, the influence of the internal layered structure and insulating medium of the target electrical equipment on the sound signal is used as the correction basis for the acoustic signal characteristic information, and the acoustic signal characteristic information is compensated and corrected specifically to obtain the abnormal acoustic fingerprint characteristic information that can accurately characterize the actual partial discharge situation inside the electrical equipment, enhancing the adaptability of the detection process to high-voltage electrical equipment with a multi-layer composite structure and filled with insulating medium inside, improving the positioning accuracy of the internal fault location of the equipment. According to the abnormal acoustic fingerprint characteristic information, abnormal component information is generated, and the corresponding visual abnormal detection report is provided to the corresponding maintenance personnel to improve the reference value of the detection result.

[0049] In some embodiments, the array acoustic signal data is acquired by a circular microphone array; the number of microphones in the circular microphone array is not less than 128; the acoustic measurement frequency range of the circular microphone array is 2 kHz - 40 kHz.

[0050] The circular microphone array can be an array structure composed of multiple microphones arranged in a circular geometric layout.

[0051] The acoustic measurement frequency range can be the frequency interval in which the ambient microphone array can effectively detect, analyze and generate an acoustic image.

[0052] Specifically, in traditional acoustic imaging technology, the traditional linear microphone array has a problem of being sensitive to directionality and is difficult to meet the sound source localization requirements under the complex three-dimensional structure of high-voltage electrical equipment. The circular array distributes microphones evenly at 360°, which can eliminate the amplitude-frequency response difference caused by the incident angle of sound waves, ensuring that sound signals in any direction can be symmetrically received. Moreover, the circular array structure can generate a three-dimensional sound field directivity map through the beamforming algorithm, significantly improving the sound source localization accuracy. According to the spatial sampling theorem, the sound source localization accuracy is positively correlated with the microphone density. Setting the number of microphones in the circular microphone array to not less than 128 can make the sound source localization accuracy reach the sub-centimeter level, and the high-density microphones can suppress environmental noise through spatial filtering, greatly improving the signal-to-noise ratio. Setting the acoustic measurement frequency range of the circular microphone array to 2 kHz - 40 kHz can effectively cover the sound signals reflected by different types of local abnormal discharges, improving the comprehensiveness of the detection process.

[0053] Through this solution, the array mode, the number of microphones, and the acoustic measurement frequency range of the microphone are constrained to meet the sound source localization requirements under the complex three-dimensional structure of high-voltage electrical equipment, improve the sound source localization accuracy, suppress environmental noise, increase the signal-to-noise ratio, and thus improve the comprehensiveness and accuracy of the detection process.

[0054] In some embodiments, according to the preset unit sound frame time, the continuous frame features of the array acoustic signal data are extracted; according to the continuous frame features, the array acoustic signal data is framed, and the time-domain features, frequency-domain features, and time-frequency joint non-stationary features of the acoustic signal are extracted; based on the time-domain features, frequency-domain features, and time-frequency joint non-stationary features of the acoustic signal, according to the mechanical vibration acoustic features and partial discharge acoustic features, the mechanical vibration acoustic signal and the partial discharge acoustic signal in the array acoustic signal data are separated; according to the partial discharge acoustic signal, the acoustic signal feature information is constructed; the mechanical vibration acoustic features are mainly characterized by low-frequency harmonic groups and show periodic attenuation characteristics; the partial discharge acoustic features are characterized by high-frequency pulse clusters and have random burst characteristics.

[0055] The preset unit sound frame time can be the preset unit time length for dividing the time frame, such as 20 ms; The continuous frame features can be the set of sound frames obtained by dividing the continuous sound signal according to the preset unit sound frame time.

[0056] The time-domain features of the acoustic signal can be the characteristic parameters that describe the statistical characteristics of the sound wave in the time dimension and are used to characterize the instantaneous energy change and periodic law of the acoustic signal.

[0057] The frequency-domain features of the acoustic signal can be the characteristic parameters that describe the statistical characteristics of the sound wave in the frequency dimension and are used to reveal the distribution law of different frequency components in the acoustic signal.

[0058] The time-frequency joint non-stationary features can be the characteristic parameters that analyze the dynamic changes of the acoustic signal by combining the time and frequency dimensions and are used to depict the transient pulse and frequency modulation phenomena in the acoustic signal.

[0059] The mechanical vibration acoustic features can be the acoustic signal features generated by the operation vibration of the mechanical components of the equipment.

[0060] The partial discharge acoustic features can be the acoustic signal features generated by local abnormal discharges inside the equipment.

[0061] The periodic attenuation characteristics can be the characteristics that the sound signal intensity shows a decaying trend within the period.

[0062] The high-frequency pulse cluster can be the high-frequency and high-energy transient pulses that continuously appear in a short time in the sound signal.

[0063] The random burst characteristics can be the characteristics that the sound signal intensity mutates in a short time.

[0064] Specifically, the acoustic signals of high-voltage electrical equipment have significant non-stationarity (such as the sudden occurrence of discharge pulses and the periodic fluctuations of mechanical vibrations). Traditional full-time analysis will obscure the characteristics of transient acoustic signals. By dividing the long-time signal into short time segments (such as 20 ms / frame), the local characteristics of different frames can be extracted to achieve frame-by-frame processing. For example, the duration of partial discharge pulses is usually in the microsecond level and only appears in a few frames. Frame-by-frame processing can accurately capture the occurrence time and spatial distribution of the partial discharge pulses, avoiding confusion with the steady-state signals of mechanical vibrations. If the vibrations caused by the operation of the equipment itself are confused with the abnormal partial discharge, it is easy to cause ineffective maintenance or delay in emergency repair due to misjudgment. The multi-channel array acoustic signals are segmented into continuous frames with a frame length of 20 ms and an overlap rate of 50%. A Hanning window is added to each frame of data to reduce spectral leakage. Each frame of data is marked with a timestamp and spatial coordinates (according to the geometric relationship of the microphone array) to construct a frame data set with spatio-temporal alignment, realizing the frame-by-frame processing of the array acoustic signal data. Furthermore, the root mean square energy, zero-crossing rate, and short-time average amplitude of each frame of signal are quantified to generate a time-domain feature vector as the time-domain feature of the acoustic signal. The fast Fourier transform is performed on each frame of acoustic signal to extract the frequency and amplitude of the harmonic components and the energy ratio of the high-frequency band (10 kHz - 40 kHz) to generate a frequency-domain feature vector as the frequency-domain feature of the acoustic signal. The multi-layer wavelet packet decomposition is performed on each frame of acoustic signal to quantify the energy entropy and instantaneous frequency variance of each sub-band, generating a time-frequency feature vector as the time-frequency joint non-stationary feature. The above time-domain feature vector, frequency-domain feature vector, and time-frequency feature vector are concatenated into a high-dimensional joint feature vector, and the high-dimensional joint feature vector is input into a pre-trained support vector machine classifier model (the training data includes labeled mechanical vibration and partial discharge acoustic signal samples, and the labeling is based on the acoustic characteristics of mechanical vibration and partial discharge). Binary classification is performed on each frame of signal to separate the mechanical vibration acoustic signal and the partial discharge acoustic signal in the array acoustic signal data, avoiding the confusion of the two signals. According to the separated partial discharge acoustic signal, the acoustic signal feature information is constructed to make the acoustic signal feature information point to the abnormal discharge situation inside the equipment.

[0065] Through this solution, the frame-by-frame processing of the array acoustic signal data is carried out, and the time-domain feature, frequency-domain feature, and time-frequency joint non-stationary feature of the acoustic signal are extracted respectively. On this basis, according to the acoustic characteristics of mechanical vibration and partial discharge, the separation between the mechanical vibration acoustic signal and the partial discharge acoustic signal in the array acoustic signal data is realized, so as to construct the acoustic signal feature information according to the partial discharge acoustic signal, make the acoustic signal feature information point to the abnormal discharge situation inside the equipment, avoid the confusion between the vibration caused by the operation of the equipment itself and the partial discharge abnormality, and prevent ineffective maintenance or delay in emergency repair due to misjudgment.

[0066] In some embodiments, the set of layered structure parameters includes information on the thickness distribution of structural layers, information on structural materials, and geometric topology information; the set of insulating medium parameters includes the type of medium, the dynamic viscosity coefficient of the medium, the acoustic velocity propagation characteristics of the medium, and the reference medium pressure.

[0067] The information on the thickness distribution of structural layers can be information describing the thickness values and their spatial distribution characteristics of each layer structure of high-voltage equipment (such as the insulating paper layer of a transformer winding, the metal layer of a GIS housing).

[0068] The information on structural materials can be information describing the types of materials (such as epoxy resin, aluminum alloy) that make up each layer structure of the equipment and their physical properties.

[0069] The geometric topology information can be information characterizing the connection relationship and shape parameters of each structural layer within the three-dimensional space of the equipment.

[0070] The type of medium can be the specific type of insulating gas inside high-voltage electrical equipment (such as SF6 gas, silicone oil).

[0071] The dynamic viscosity coefficient of the medium can be a quantitative index describing the flow resistance characteristics of the insulating medium.

[0072] The acoustic velocity propagation characteristics of the medium can be the propagation velocity of sound waves in the insulating medium.

[0073] The reference medium pressure can be the standard working pressure of the insulating medium inside high-voltage electrical equipment during operation.

[0074] Specifically, the interior of high-voltage electrical equipment (such as enclosed gas-insulated switches and transformers) is composed of multiple layers of heterogeneous materials. During the propagation of sound waves, multiple reflections, transmissions, and attenuations will occur. When the partial discharge acoustic signal passes through the layered structure inside the high-voltage electrical equipment, the degree of its energy attenuation is directly related to the characteristics of the structural materials. The acoustic impedance differences between different materials (such as metals and polymers) are significant. On this basis, the propagation of the acoustic signal is also affected by the thickness and geometric shape (such as cavities) of the corresponding materials. By collecting information on the thickness distribution of structural layers, information on structural materials, and geometric topology information, the influence of the internal layered structure of high-voltage electrical equipment on the partial discharge acoustic signal can be comprehensively reflected; the insulating medium is filled between different hierarchical structures inside the high-voltage electrical equipment and is on the propagation path of the partial discharge acoustic signal. There are significant differences in the sound propagation speeds in insulating media with different characteristics. The sound absorption mechanisms of gases (such as SF6) and liquids (such as transformer oil) are significantly different. The viscosity of the insulating medium will cause a phase delay during the propagation of sound waves, and the change in the medium pressure will change the acoustic impedance and attenuation coefficient. By clarifying the type of medium, the dynamic viscosity coefficient of the medium, the acoustic velocity propagation characteristics of the medium, and the reference medium pressure, a reliable data basis can be provided for comprehensively analyzing the influence of the insulating medium on the acoustic signal.

[0075] Through this scheme, a layered structure parameter set is constructed according to the structural layer thickness distribution information, structural material information and geometric topology information. The influence of the multi-layer structure inside the high-voltage electrical equipment on the partial discharge acoustic signal is reflected from the three dimensions of material properties, material thickness and geometric shape. At the same time, an insulating medium parameter set is constructed according to the medium type, medium dynamic viscosity coefficient, medium sound speed propagation characteristics and medium pressure reference, providing a reliable data basis for a comprehensive analysis of the influence of insulating media on acoustic signals.

[0076] In some embodiments, according to the layered structure parameter set, a physical field simulation model based on multi-layer geometric structure sound propagation is constructed; according to the physical field simulation model, the reflection-transmission attenuation coefficient of the local discharge acoustic signal in different layered structures is determined; according to the insulating medium parameter set, the propagation path of the local discharge acoustic signal is phase compensated and corrected to determine the dielectric absorption coefficient; according to the reflection-transmission attenuation coefficient and the dielectric absorption coefficient, the local discharge acoustic signal is targetedly compensated and corrected to determine the abnormal soundprint feature information.

[0077] The physical field simulation model can be a numerical model of multi-layer medium sound propagation built based on the finite element method, integrating structural geometry, material properties and boundary conditions, and is used to fit the propagation laws of sound waves in complex equipment.

[0078] The reflection-transmission attenuation coefficient can be a quantitative indicator to quantify the energy loss of sound waves at the layered interface.

[0079] Phase compensation correction can be a sound signal correction and repair process for the sound wave phase distortion caused by medium viscosity and inhomogeneity.

[0080] The medium absorption coefficient can be a normalized quantitative indicator that characterizes the intensity of the medium's absorption of sound wave energy.

[0081] Specifically, import the layered structure parameter set into a physical field modeling tool, such as the acoustic module of COMSOL Multiphysics. Divide the adaptive grid according to the geometric topology information, load the acoustic impedance parameters corresponding to the structural material information at the interface of each layer structure, construct a physical field simulation model, and simulate the propagation state of the partial discharge acoustic signal in different layered structures through the physical field simulation model, and quantify the reflection-transmission attenuation coefficient used to characterize the attenuation effect suffered by the partial discharge acoustic signal during propagation in the layered structure; according to the properties of different medium types, input the insulating medium parameter set into the corresponding model (for example, the Stokes model in fluid mechanics is used for gas insulating media, and the Wood model is used for liquid insulating media), simulate the propagation path characteristics of the partial discharge acoustic signal in the insulating medium, and determine the medium absorption coefficient reflecting the influence of the insulating medium on the propagation of the acoustic signal; perform time-frequency decomposition on the partial discharge acoustic signal, and reverse correct the phase of the acoustic signal according to the reflection-transmission attenuation coefficient and the medium absorption coefficient corresponding to the propagation path, reconstruct the pure discharge acoustic fingerprint, and compare the pure discharge acoustic fingerprint with the acoustic fingerprint information obtained in the experiment under the condition of no partial discharge abnormality to extract the abnormal acoustic fingerprint characteristic information.

[0082] Through this solution, according to the layered structure parameter set, construct a physical field simulation model to simulate the negative impact suffered by the acoustic signal during the propagation process between different layered structures inside the high-voltage electrical equipment, fit the corresponding reflection-transmission attenuation coefficient, and according to the insulating medium parameter set, analyze the propagation characteristics of the acoustic signal in the insulating medium, fit the medium absorption coefficient reflecting the influence of the insulating medium on the propagation of the acoustic signal, and characterize the attenuation characteristics of the acoustic signal under the dual influence through the reflection-transmission attenuation coefficient and the medium absorption coefficient, and thus inversely deduce the actual sound signal characteristics caused by partial discharge, accurately locate the abnormal acoustic fingerprint characteristic information, and improve the judgment accuracy for local abnormal discharge.

[0083] In some embodiments, according to the structural layer thickness distribution information and geometric topology information, construct a three-dimensional space grid model, parametrically characterize the geometric characteristics of each layer structure, and determine several hierarchical grid parameter nodes; based on the structural material information, load the acoustic impedance parameters of the corresponding structural material at each hierarchical grid parameter node, and according to the acoustic impedance parameters, simulate and mark the time delay characteristics, attenuation characteristics and superposition characteristics of the partial discharge acoustic signal in each propagation path through the ray tracing algorithm to construct a physical field simulation model.

[0084] The three-dimensional space grid model can be a numerical model that discretizes the device geometric structure into finite element units and is used for acoustic field simulation.

[0085] The hierarchical grid parameter nodes can be key nodes that mark the junctions of each structural layer in the grid model and are used to load material properties and acoustic boundary conditions.

[0086] The acoustic impedance parameter can be the product of the material density and the sound velocity, which characterizes the reflection / transmission characteristics of sound waves at the medium interface.

[0087] The ray tracing algorithm can be an algorithm that quantifies the reflection, transmission, and attenuation processes of sound waves in a complex structure by simulating the propagation path of sound wave rays. The ray tracing algorithm can be implemented using the high-frequency approximation algorithm module in professional acoustic simulation software (such as PZFlex or Actran).

[0088] The time delay feature can be the difference in the propagation time of sound waves from the discharge source to the microphone array.

[0089] The attenuation feature can be the energy loss rate of sound waves caused by medium absorption and interface reflection during propagation.

[0090] The superposition feature can be the superposition interference effect of multi-path sound waves (such as reflected waves and direct waves) at the receiving point.

[0091] Specifically, based on the structural layer thickness distribution information and geometric topology information, unstructured mesh division is performed using high-order tetrahedral elements to ensure the mesh quality in areas with complex curvature, generating a three-dimensional space mesh model containing hierarchical mesh parameter nodes. According to the structural material information associated with the hierarchical mesh parameter nodes, the acoustic impedance parameters are marked at the corresponding hierarchical mesh parameter nodes. By randomly generating partial discharge points and using these as starting points, the time delay feature, attenuation feature, and superposition feature of the partial discharge acoustic signals emitted from the partial discharge points in each propagation path are simulated and statistically analyzed through the ray tracing algorithm, constructing a physical field simulation model for characterizing the propagation characteristics of partial discharge acoustic signals.

[0092] Through this solution, according to the structural layer thickness distribution information and geometric topology information, a three-dimensional space mesh model containing hierarchical mesh parameter nodes is constructed to realize the acoustic field simulation of the three-dimensional irregular structure inside high-voltage electrical equipment, eliminating the source location offset caused by two-dimensional models. On this basis, based on the acoustic impedance parameters reflected in the structural material information, the time delay feature, attenuation feature, and superposition feature of the partial discharge acoustic signals in each propagation path are simulated and marked through the ray tracing algorithm, constructing a physical field simulation model, realizing the high-precision simulation of the sound wave propagation characteristics under the complex structure of high-voltage equipment, providing a reliable numerical basis for the physical field compensation of abnormal acoustic fingerprint features, and significantly improving the location accuracy and diagnostic reliability of partial discharge detection.

[0093] In some embodiments, a basic dynamic propagation velocity field model of partial discharge acoustic signals in an insulating medium is constructed according to the medium type and the acoustic velocity propagation characteristics of the medium; based on the basic dynamic propagation velocity field model, the viscous phase delay of the partial discharge acoustic signal in the insulating medium is simulated according to the dynamic viscosity coefficient of the medium and the reference medium pressure; the viscous phase delay is positively correlated with the dynamic viscosity coefficient of the medium; the viscous phase delay is negatively correlated with the reference medium pressure; the viscous phase delay is normalized to determine the medium absorption coefficient.

[0094] The basic dynamic propagation velocity field model can be a numerical model describing the spatial distribution of the propagation velocity of sound waves in a non-uniform medium.

[0095] The viscous phase delay acoustic wave can be the propagation time delay caused by the viscous resistance of the medium.

[0096] Specifically, a sound velocity calculation model is selected according to the medium type: for gas media, based on the modified ideal gas equation, and for liquid media, based on the Tait equation of state. The experimental data of the acoustic velocity propagation characteristics of the medium (obtained through sound propagation experiments under different insulating media) are imported, and the model parameters are fitted by the least squares method to construct the basic dynamic propagation velocity field model; based on the basic dynamic propagation velocity field model, the dynamic viscosity coefficient of the medium and the reference medium pressure are introduced to simulate the viscous phase delay of the partial discharge acoustic signal under the influence of the viscosity and pressure of the insulating medium. Specifically, the greater the viscosity of the insulating medium, the greater the corresponding viscous phase delay. Conversion is performed according to the positive correlation coefficient between the viscosity of the insulating medium and the viscous phase delay. The greater the pressure of the insulating medium, the lower the corresponding viscous phase delay. Conversion is performed according to the negative correlation coefficient between the pressure of the insulating medium and the viscous phase delay. After the above positive and negative adjustments, the viscous phase delay of the partial discharge acoustic signal in the insulating medium is obtained, which characterizes the intensity of the absorption effect suffered by the sound signal when passing through the current insulating medium. Further, the minimum-maximum normalization algorithm is used to normalize the viscous phase delay, and the quantified medium absorption coefficient is mapped to the numerical interval of (0,1).

[0097] Through this solution, according to the medium type and the acoustic velocity propagation characteristics of the medium, a basic dynamic propagation velocity field model is constructed to reflect the basic propagation characteristics of the sound signal in the current insulating medium. On this basis, the dynamic viscosity coefficient of the medium and the reference medium pressure are introduced to simulate the viscous phase delay generated under the dual influence of the viscosity and pressure of the insulating medium on the partial discharge acoustic signal, and the viscous phase delay is normalized to obtain the medium absorption coefficient, so as to accurately reflect the actual propagation state of the sound signal in the insulating medium and improve the accuracy of subsequent positioning analysis for the partial discharge location.

[0098] In some embodiments, after determining several hierarchical grid parameter nodes in the physical field simulation model, the reflection-transmission attenuation coefficient and the medium absorption coefficient are substituted into an inverse ray tracing algorithm to invert the sound source propagation path, determine the coordinate point of the sound source of the local discharge sound signal in the three-dimensional space grid model, and generate the three-dimensional coordinates of the abnormal position; based on the geometric topology information, the relative position of the abnormal equipment component and the local discharge is determined according to the three-dimensional coordinates of the abnormal position; according to the relative position of the abnormal equipment component and the local discharge, a preset equipment abnormality information library is retrieved to determine abnormality troubleshooting suggestions; according to the abnormal equipment component, the relative position of the local discharge and the abnormality troubleshooting suggestions, the abnormal component information is constructed.

[0099] The inverse ray tracing algorithm may be an algorithm for tracing the sound wave propagation path from the microphone array receiving point in reverse to the partial discharge location.

[0100] Inverting the sound source propagation path can be a process of reversely deducing the sound signal propagation path to locate the sound source.

[0101] The abnormal position three-dimensional coordinates may be spatial coordinates for locating a local discharge source in a three-dimensional model of the device.

[0102] The abnormal equipment component may be the specific part where the partial discharge occurs.

[0103] The relative position of the partial discharge may be the relative position of a specific discharge position in a corresponding abnormal equipment component.

[0104] The preset equipment abnormality information base may be a knowledge base storing historical defect cases, including typical discharge locations, fault types, and corresponding abnormality troubleshooting suggestions.

[0105] The abnormality troubleshooting suggestions may be maintenance measures recommended for the current abnormality characteristics.

[0106] Specifically, taking the current position of the microphone array as a reference, the reflection-transmission attenuation coefficient and the dielectric absorption coefficient are loaded in the physical field simulation model, and the inverse ray tracing algorithm is run. According to the characteristics of the currently collected sound signal, the sound source propagation path is reversely deduced, and the sound source corresponding to the current sound signal inside the high-voltage electrical equipment is located, and then the coordinate point of the sound source of the local discharge sound signal in the three-dimensional space grid model is determined, and the three-dimensional coordinates of the abnormal position are generated. The equipment component at the three-dimensional coordinates of the abnormal position is the abnormal equipment component. Taking the space grid corresponding to the abnormal equipment component as a reference, a new space coordinate system is established to determine the relative position of the local discharge. According to the abnormal equipment component and the relative position of the local discharge, the preset equipment abnormality information database is retrieved to determine the abnormal troubleshooting suggestions that can be used by maintenance personnel when the current abnormal equipment component has abnormal discharge at the current relative position of the local discharge. By integrating the abnormal equipment component, the relative position of the local discharge and the abnormal troubleshooting suggestions, the abnormal component information is constructed.

[0107] Through this solution, based on the physical field simulation model, according to the reflection-transmission attenuation coefficient and the medium absorption coefficient, using the inverse ray tracing algorithm, the propagation path of the partial discharge acoustic signal currently collected is deduced, and then the sound source of the partial discharge acoustic signal inside the high-voltage electrical equipment is located, the three-dimensional coordinates of the abnormal position are obtained, combined with the geometric topology information, the abnormal equipment components and the relative position of the partial discharge are determined, and the preset equipment abnormal information database is retrieved to determine the abnormal troubleshooting suggestions. Integrating according to the abnormal equipment components, the relative position of the partial discharge and the abnormal troubleshooting suggestions, constructing abnormal component information, realizing the accurate positioning of the abnormal discharge position inside the high-voltage electrical equipment, and providing corresponding abnormal troubleshooting suggestions, effectively improving the maintenance efficiency and effect of the subsequent maintenance personnel.

[0108] In some embodiments, obtain the maintenance personnel's maintenance feedback information, analyze the maintenance personnel's maintenance feedback information, and determine whether the maintenance personnel adopt the abnormal troubleshooting suggestions; if the maintenance personnel adopt the corresponding abnormal troubleshooting suggestions, then increase the influence factor of the abnormal troubleshooting suggestions; if the maintenance personnel do not adopt the corresponding abnormal troubleshooting suggestions, then incorporate the troubleshooting process in the maintenance personnel's maintenance feedback information into the troubleshooting suggestion list corresponding to the relative position of the partial discharge under the corresponding abnormal equipment components in the preset equipment abnormal information database, and set its corresponding influence factor to the preset influence reference value; each abnormal troubleshooting suggestion in the troubleshooting suggestion list is sorted according to the numerical value of its corresponding influence factor.

[0109] The maintenance personnel's maintenance feedback information can be text records or structured form data submitted by the maintenance personnel through a mobile terminal after on-site maintenance.

[0110] The influence factor can be a weight value that quantitatively characterizes the historical effectiveness of the troubleshooting suggestions.

[0111] The troubleshooting suggestion list can be a set of maintenance plans sorted by influence factors generated for the current partial discharge abnormality.

[0112] The preset influence reference value can be the initial influence factor of a newly added unverified suggestion.

[0113] Specifically, due to the coupling effect of many different influencing factors on the partial discharge abnormality of high-voltage electrical equipment, the abnormal troubleshooting suggestions constructed based on historical experience and expert experience cannot fully cover the actual abnormal situations. Based on the collected maintenance personnel's maintenance feedback information, using the adjustment mechanism of the influence factor, the abnormal troubleshooting suggestions in the abnormal troubleshooting suggestion list are dynamically updated and the recommendation priorities are adjusted, establishing a complete closed loop of "suggestion generation - on-site verification - knowledge evolution", enabling the abnormal troubleshooting system to have the ability of continuous self-optimization to adapt to complex actual situations and further improving the maintenance efficiency and effect of the maintenance personnel.

[0114] Through this solution, based on the maintenance feedback information collected, using the adjustment mechanism of influence factors, the anomaly detection suggestions in the anomaly detection suggestion list are dynamically updated and the recommended priorities are adjusted, establishing a complete closed-loop of "suggestion generation - on-site verification - knowledge evolution", enabling the anomaly detection system to have the ability of continuous self-optimization to adapt to complex actual situations, and further improving the maintenance efficiency and maintenance effect of maintenance personnel.

[0115] Figure 3 The following is a schematic structural diagram of a high-voltage electrical equipment inspection system based on microphone array measurement provided by an embodiment of the present application. Figure 3 As shown, a high-voltage electrical equipment inspection system 300 based on microphone array measurement in this embodiment includes: a feature analysis module 301, a feature correction module 302, and an anomaly location module 303.

[0116] The feature analysis module 301 is configured to obtain array acoustic signal data, analyze the array acoustic signal data, and determine acoustic signal feature information. The feature correction module 302 is configured to obtain a layered structure parameter set and an insulating medium parameter set of the target electrical equipment, and perform targeted compensation and correction on the acoustic signal feature information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal acoustic pattern feature information. The anomaly location module 303 is configured to generate abnormal component information according to the abnormal acoustic pattern feature information, and determine and output an anomaly detection report.

[0117] Optionally, the array acoustic signal data in the feature analysis module 301 is obtained through a circular microphone array. The number of microphones in the circular microphone array is not less than 128. The acoustic measurement frequency range of the circular microphone array is 2 kHz - 40 kHz.

[0118] Optionally, the feature analysis module 301 is specifically configured to: Perform frame division processing on the array acoustic signal data according to the continuous frame features of the array acoustic signal data, and extract acoustic signal time-domain features, acoustic signal frequency-domain features, and time-frequency joint non-stationary features. Based on the acoustic signal time-domain features, the acoustic signal frequency-domain features, and the time-frequency joint non-stationary features, separate the mechanical vibration acoustic signals and partial discharge acoustic signals in the array acoustic signal data according to mechanical vibration acoustic features and partial discharge acoustic features. Construct the acoustic signal feature information according to the partial discharge acoustic signals. The mechanical vibration acoustic features are mainly low-frequency harmonic groups and show periodic attenuation characteristics. The local discharge acoustic characteristics are manifested as high-frequency pulse clusters and have random burst characteristics.

[0119] Optionally, the layered structure parameter set in the feature correction module 302 includes structure layer thickness distribution information, structure material information and geometric topology information; The insulating medium parameter set includes medium type, medium dynamic viscosity coefficient, medium sound velocity propagation characteristics and medium pressure reference.

[0120] Optionally, the feature correction module 302 is specifically used to: According to the layered structure parameter set, a physical field simulation model based on multi-layer geometric structure sound propagation is constructed; Determining the reflection-transmission attenuation coefficient of the partial discharge acoustic signal in different layered structures according to the physical field simulation model; According to the insulating medium parameter set, performing phase compensation correction on the propagation path of the partial discharge acoustic signal to determine the dielectric absorption coefficient; According to the reflection-transmission attenuation coefficient and the medium absorption coefficient, targeted compensation and correction are performed on the partial discharge acoustic signal to determine the abnormal soundprint feature information.

[0121] Optionally, when constructing a physical field simulation model based on multi-layer geometric structure sound propagation according to the layered structure parameter set, the feature correction module 302 is specifically used to: According to the thickness distribution information of the structural layer and the geometric topology information, a three-dimensional space grid model is constructed, the geometric features of each layer of the structure are parameterized, and a plurality of hierarchical grid parameter nodes are determined; Based on the structural material information, the acoustic impedance parameters of the corresponding structural material are loaded at each of the hierarchical grid parameter nodes, and according to the acoustic impedance parameters, the delay characteristics, attenuation characteristics and superposition characteristics of the local discharge acoustic signal in each propagation path are simulated and marked by a ray tracing algorithm to construct the physical field simulation model.

[0122] Optionally, the feature correction module 302 performs phase compensation correction on the propagation path of the partial discharge acoustic signal according to the insulating medium parameter set to determine the medium absorption coefficient, specifically for: Constructing a basic dynamic propagation velocity field model of the partial discharge acoustic signal in the insulating medium according to the medium type and the sound velocity propagation characteristics of the medium; Based on the basic dynamic propagation velocity field model, according to the medium dynamic viscosity coefficient and the medium pressure reference, simulating the viscous phase delay of the partial discharge acoustic signal in the insulating medium; The viscous phase delay is positively correlated with the dynamic viscosity coefficient of the medium; The viscous phase delay is negatively correlated with the medium pressure reference; Normalize the viscous phase delay to determine the medium absorption coefficient.

[0123] Optionally, the anomaly location module 303 is specifically configured to: After determining several hierarchical grid parameter nodes in the physical field simulation model, substitute the reflection-transmission attenuation coefficient and the medium absorption coefficient into the inverse ray tracing algorithm, invert the sound source propagation path, determine the coordinate points of the sound source of the partial discharge acoustic signal in the three-dimensional space grid model, and generate three-dimensional coordinates of the anomaly location; Based on the geometric topology information, determine the abnormal equipment component and the relative position of the partial discharge according to the three-dimensional coordinates of the anomaly location; According to the abnormal equipment component and the relative position of the partial discharge, retrieve the preset equipment anomaly information library to determine the anomaly troubleshooting suggestions; Construct the abnormal component information according to the abnormal equipment component, the relative position of the partial discharge, and the anomaly troubleshooting suggestions.

[0124] Optionally, the high-voltage electrical equipment inspection system 300 based on microphone array measurement further includes a suggestion adjustment module 304, which is used for: Obtain the maintenance personnel's inspection feedback information, analyze the maintenance personnel's inspection feedback information, and determine whether the maintenance personnel adopt the anomaly troubleshooting suggestions; If the maintenance personnel adopt the corresponding anomaly troubleshooting suggestions, increase the influence factor of the anomaly troubleshooting suggestions; If the maintenance personnel do not adopt the corresponding anomaly troubleshooting suggestions, incorporate the troubleshooting process in the maintenance personnel's inspection feedback information into the troubleshooting suggestion list corresponding to the relative position of the partial discharge under the corresponding abnormal equipment component in the preset equipment anomaly information library, and set its corresponding influence factor to the preset influence reference value; The various anomaly troubleshooting suggestions in the troubleshooting suggestion list are sorted according to the numerical values of their corresponding influence factors.

[0125] The system of this embodiment can be used to execute the method of any of the above embodiments, and its implementation principle and technical effects are similar, which will not be elaborated here.

Claims

1. A patrol inspection method for high-voltage electrical equipment based on microphone array measurement, characterized in that, include: Acquire array acoustic signal data of target electrical equipment, analyze the array acoustic signal data, and determine acoustic signal characteristic information; The acoustic signal feature information is a feature set extracted from the array acoustic signal data and used to characterize the operating state of the target device; Acquire a layered structure parameter set and an insulating medium parameter set of the target electrical device, and perform targeted compensation and correction on the acoustic signal characteristic information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal voiceprint characteristic information; The abnormal voiceprint feature information is the sound feature information used to map the local abnormal discharge situation inside the target electrical equipment; Based on the abnormal voiceprint feature information, abnormal component information is generated, and an abnormal detection report is determined and output.

2. The method according to claim 1, characterized in that The array acoustic signal data is obtained through a ring microphone array; The number of microphones in the annular microphone array is not less than 128; The acoustic measurement frequency range of the annular microphone array is 2 kHz-40 kHz.

3. The method according to claim 2, characterized in that The step of analyzing the array acoustic signal data to determine acoustic signal characteristic information includes: Extracting continuous frame features of the array sound signal data according to a preset unit sound frame time; According to the continuous frame features, the array acoustic signal data is framed to extract the acoustic signal time domain features, acoustic signal frequency domain features and time-frequency joint non-stationary features; Based on the time domain features of the acoustic signal, the frequency domain features of the acoustic signal and the time-frequency joint non-stationary features, and according to the mechanical vibration acoustic features and the partial discharge acoustic features, the mechanical vibration acoustic signal and the partial discharge acoustic signal in the array acoustic signal data are separated; Constructing characteristic information of the acoustic signal according to the partial discharge acoustic signal; The mechanical vibration acoustic characteristics are characterized by being dominated by low-frequency harmonic groups and exhibiting periodic attenuation characteristics; The local discharge acoustic characteristics are manifested as high-frequency pulse clusters and have random burst characteristics.

4. The method according to claim 3, characterized in that, The layered structure parameter set includes structure layer thickness distribution information, structure material information and geometric topology information; The insulating medium parameter set includes medium type, medium dynamic viscosity coefficient, medium sound velocity propagation characteristics and medium pressure reference.

5. The method according to claim 4, characterized in that The method of performing targeted compensation and correction on the acoustic signal characteristic information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal voiceprint characteristic information includes: According to the layered structure parameter set, a physical field simulation model based on multi-layer geometric structure sound propagation is constructed; Determining the reflection-transmission attenuation coefficient of the partial discharge acoustic signal in different layered structures according to the physical field simulation model; According to the insulating medium parameter set, performing phase compensation correction on the propagation path of the partial discharge acoustic signal to determine the dielectric absorption coefficient; According to the reflection-transmission attenuation coefficient and the medium absorption coefficient, targeted compensation and correction are performed on the partial discharge acoustic signal to determine the abnormal soundprint feature information.

6. The method according to claim 5, wherein The method of constructing a physical field simulation model based on multi-layer geometric structure sound propagation according to the layered structure parameter set includes: According to the thickness distribution information of the structural layer and the geometric topology information, a three-dimensional space grid model is constructed, the geometric features of each layer of the structure are parameterized, and a plurality of hierarchical grid parameter nodes are determined; Based on the structural material information, the acoustic impedance parameters of the corresponding structural material are loaded at each of the hierarchical grid parameter nodes, and according to the acoustic impedance parameters, the delay characteristics, attenuation characteristics and superposition characteristics of the local discharge acoustic signal in each propagation path are simulated and marked by a ray tracing algorithm to construct the physical field simulation model.

7. The method according to claim 5, characterized in that, The step of performing phase compensation correction on the propagation path of the partial discharge acoustic signal according to the insulating medium parameter set to determine the dielectric absorption coefficient includes: Constructing a basic dynamic propagation velocity field model of the partial discharge acoustic signal in the insulating medium according to the medium type and the sound velocity propagation characteristics of the medium; Based on the basic dynamic propagation velocity field model, according to the medium dynamic viscosity coefficient and the medium pressure reference, simulating the viscous phase delay of the partial discharge acoustic signal in the insulating medium; The viscous phase delay is positively correlated with the dynamic viscosity coefficient of the medium; The viscous phase delay is negatively correlated with the medium pressure reference; The viscous phase delay is normalized to determine the medium absorption coefficient.

8. The method according to claim 7, characterized in that The generating abnormal component information according to the abnormal voiceprint feature information includes: After determining several levels of grid parameter nodes in the physical field simulation model, the reflection-transmission attenuation coefficient and the medium absorption coefficient are substituted into an inverse ray tracing algorithm to invert the sound source propagation path, determine the coordinate point of the sound source of the partial discharge sound signal in the three-dimensional space grid model, and generate the three-dimensional coordinates of the abnormal position; Based on the geometric topology information, determining the relative position of the abnormal equipment component and the partial discharge according to the three-dimensional coordinates of the abnormal position; According to the relative position of the abnormal equipment component and the partial discharge, a preset equipment abnormality information database is retrieved to determine abnormality troubleshooting suggestions; The abnormal component information is constructed based on the abnormal equipment component, the relative position of the partial discharge and the abnormality troubleshooting suggestion.

9. The method according to claim 8, characterized in that, The method further comprises: Obtain maintenance feedback information from maintenance personnel, analyze the maintenance feedback information from maintenance personnel, and determine whether the maintenance personnel adopt the abnormality troubleshooting suggestions; If the maintenance personnel adopt the corresponding abnormality troubleshooting suggestion, then the impact factor of the abnormality troubleshooting suggestion is increased; If the maintenance personnel do not adopt the corresponding abnormality troubleshooting suggestion, the troubleshooting process in the maintenance personnel's maintenance feedback information is included in the troubleshooting suggestion list corresponding to the relative position of the partial discharge under the abnormal equipment component in the preset equipment abnormality information library, and the corresponding impact factor is set as the preset impact reference value; The abnormality troubleshooting suggestions in the troubleshooting suggestion list are sorted according to the numerical values ​​of their corresponding impact factors.

10. A high-voltage electrical equipment inspection system based on microphone array measurement, characterized in that, include: A feature analysis module, used to obtain array acoustic signal data, analyze the array acoustic signal data, and determine acoustic signal feature information; A feature correction module, configured to obtain a layered structure parameter set and an insulating medium parameter set of a target electrical device, and perform targeted compensation and correction on the acoustic signal feature information according to the layered structure parameter set and the insulating medium parameter set to determine abnormal acoustic feature information; An abnormal positioning module, configured to generate abnormal component information according to the abnormal acoustic feature information, and determine and output an abnormal detection report.

Images