GIL particle defect online monitoring system and method based on intelligent sensing system

Abstract

The invention discloses a GIL particle defect on-line monitoring system and method based on an intelligent sensing system, and relates to the technical field of power monitoring, the GIL particle defect on-line monitoring system comprises a defect monitoring platform, the defect monitoring platform is in communication connection with an intelligent sensor network module, a signal conditioning module, a feature extraction module, a defect identification positioning module and a monitoring early warning module, wherein the modules are connected through electric signals; and the intelligent sensor network module is used for deploying a plurality of intelligent ultrasonic sensors on the GIL to form a sensor network. A plurality of intelligent ultrasonic sensors are deployed on a GIL to form a sensor network, ultrasonic propagation speed errors are corrected in combination with an intelligent algorithm, defect positions are accurately calculated by using a triangulation positioning method, and sensor position information and signal propagation characteristics are comprehensively considered; the problem of inaccurate positioning caused by propagation velocity uncertainty and sensor installation deviation in a traditional method is effectively solved, and the defect positioning precision is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power monitoring, and in particular to a GIL particle defect online monitoring system and method based on an intelligent sensing system. Background Art

[0002] As a highly efficient power transmission device, GIL (Gas Insulated Transmission Line) is widely used in high-voltage, large-capacity transmission scenarios. However, GILs may contain internal particle defects. These particles, under the action of the electric field, may cause partial discharge, thereby affecting the insulation performance and operational reliability of the GIL. Therefore, online monitoring of GIL particle defects is necessary. Online monitoring of GIL particle defects typically includes ultra-high frequency (UHF) detection, ultrasonic detection, and optical detection methods to ensure real-time monitoring of particle defects within the GIL.

[0003] In the prior art, when ultrasonic methods are used to monitor GIL particle defects, the uncertainty of the propagation speed of ultrasonic waves inside the GIL and the deviation of the sensor installation position and angle will lead to low positioning accuracy, making it difficult to accurately determine the specific location of the defect. In addition, the ultrasonic signal characteristics generated by different types of particle defects are similar, which will further increase the defect identification accuracy. Therefore, how to improve the accuracy of defect positioning and identification through a sensor network formed by multiple intelligent ultrasonic sensors combined with the characteristic parameters of the ultrasonic signal is a problem to be solved by the present invention. To this end, a GIL particle defect online monitoring system and method based on an intelligent sensing system are proposed. Summary of the Invention

[0004] The present invention aims to provide a system and method for online monitoring of GIL particle defects based on an intelligent sensing system to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] In a first aspect, an online monitoring system for GIL particle defects based on an intelligent sensing system includes a defect monitoring platform, wherein the defect monitoring platform is communicatively connected to an intelligent sensor network module, a signal conditioning module, a feature extraction module, a defect identification and positioning module, and a monitoring and early warning module, wherein the modules are electrically connected;

[0007] The intelligent sensor network module is used to deploy multiple intelligent ultrasonic sensors on the GIL to form a sensor network to collect the original ultrasonic signals inside the GIL, cover the monitoring area, reduce the impact of signal attenuation, and improve the comprehensiveness of signal acquisition;

[0008] The signal conditioning module is used to perform pre-processing operations including filtering and amplification on the collected original ultrasonic signals to suppress noise and enhance effective signals;

[0009] The feature extraction module is used to extract characteristic parameters representing the characteristics of the defect signal from the preprocessed ultrasonic signal, and determine the baseline value of each characteristic parameter to form a defect feature sequence;

[0010] The defect recognition and positioning module is used to combine the defect feature sequence and the pre-built defect recognition model to identify the particle defects and locate the position of the defects to improve the positioning accuracy;

[0011] The monitoring and early warning module is used to display monitoring data and analysis results, issue early warnings for identified and located defects, and provide maintenance decision suggestions.

[0012] A further improvement of the technical solution of the present invention is that the intelligent sensor network module specifically includes:

[0013] Determine the deployment location of intelligent ultrasonic sensors based on the GIL's structural layout, length, insulator distribution, and operating environment through a combination of computer simulation and on-site investigation.

[0014] Fix the smart ultrasonic sensors to the GIL housing according to the planned locations. During installation, ensure good contact between the sensors and the GIL housing to reduce interference and attenuation during signal propagation. After installation, debug each smart ultrasonic sensor individually, including checking key parameters such as sensor sensitivity, frequency response range, and directionality. Also, calibrate the receiving direction and angle of the smart ultrasonic sensor to enable it to accurately capture the ultrasonic signal inside the GIL. Connect each smart ultrasonic sensor wirelessly to form a smart sensor network.

[0015] All installed intelligent ultrasonic sensors are connected to the communication network of the monitoring system, and each intelligent ultrasonic sensor is used to collect the original ultrasonic signals inside the GIL in real time to achieve collaborative work between sensors.

[0016] A further improvement of the technical solution of the present invention is that the signal conditioning module specifically includes:

[0017] The collected raw ultrasonic signal is preliminarily filtered and processed. At the same time, the analog filtering circuit is used to further reduce random noise and background interference, so that the ultrasonic signal is initially focused within the target frequency range.

[0018] Amplify the ultrasonic signal after preliminary filtering, select the corresponding amplification factor according to the initial intensity of the ultrasonic signal, so that the signal amplitude is increased to a level suitable for subsequent processing, and perform linear compensation on the amplified ultrasonic signal;

[0019] The amplified signal is finely filtered to further remove residual noise components. At the same time, the filtered signal is smoothed to remove existing burrs and abnormal fluctuations, ultimately optimizing the signal quality.

[0020] A further improvement of the technical solution of the present invention is that the feature extraction module specifically includes:

[0021] Selecting characteristic parameters for characterizing particle defect characteristics from the preprocessed ultrasonic signal, wherein the characteristic parameters selected include signal amplitude, signal frequency, signal duration, number of discharges, signal rise time, and entropy of the signal time-frequency distribution based on the characteristics of the GIL particle defect signal;

[0022] The pre-processed ultrasonic signal is analyzed to extract selected characteristic parameters. The signal amplitude, signal duration, and signal rise time are obtained through time domain analysis. The signal frequency is determined by frequency domain analysis. The number of discharges is counted by pulse counting. The entropy of the signal time-frequency distribution is calculated using wavelet transform tools. During the extraction process, the accuracy and stability of each characteristic parameter are ensured to avoid extraction errors caused by signal fluctuations or noise interference.

[0023] Based on the extracted characteristic parameters, the baseline value of each characteristic parameter is determined, that is, the reference value under normal operating conditions. Through comparative analysis, the extraction results of each characteristic parameter are arranged in chronological order to form a defect feature sequence. Among them, the setting of the baseline value is based on the ultrasonic signal characteristics of the GIL during normal operation and is obtained through long-term monitoring data statistics.

[0024] A further improvement of the technical solution of the present invention is that: the defect recognition and positioning module includes a defect pattern recognition unit and a defect positioning unit;

[0025] The defect pattern recognition unit is used to construct a defect recognition model using a machine learning algorithm, analyze characteristic parameters, and identify particle defects;

[0026] The defect locating unit is used to correct the propagation velocity error and locate the specific position of the defect based on the identified particle defect and in combination with the position information and characteristic parameters of the intelligent ultrasonic sensor.

[0027] A further improvement of the technical solution of the present invention is that the defect pattern recognition unit specifically includes:

[0028] Collect a large number of GIL particle defect samples of known types, covering both normal and defective conditions, extract their characteristic parameters and label the defect types, construct a comprehensive dataset, and divide it into training and test sets;

[0029] A machine learning algorithm based on a support vector machine model was selected as the infrastructure for building the defect recognition model. The support vector machine model was trained using a training set. The model's performance was optimized by adjusting the algorithm's parameters, improving the accuracy and robustness of classification and recognition. The characteristic patterns of different types of particle defects were learned. The trained model was tested using a test set, and metrics such as the model's classification accuracy, recall rate, and F1 score were evaluated. Based on the evaluation results, the model's strengths and weaknesses were analyzed, and the model was further optimized to improve its generalization and defect recognition capabilities, ultimately resulting in a final defect recognition model.

[0030] The defect recognition model is applied to actual GIL particle defect monitoring, and the extracted characteristic parameters are classified and identified in real time to determine whether there are defects and determine the defect type that matches the characteristic parameters.

[0031] A further improvement of the technical solution of the present invention is that the defect location unit specifically includes:

[0032] After identifying the particle defects, the defect location unit extracts the position information provided by the intelligent ultrasonic sensors and determines the specific coordinates of each intelligent ultrasonic sensor in the GIL. At the same time, based on the theoretical propagation speed formula of ultrasonic waves in the GIL medium, combined with the initial conditions of the equipment material and ambient temperature, it preliminarily calculates the propagation speed of ultrasonic waves in the GIL.

[0033] According to the propagation characteristics of ultrasonic signals in GIL and the actual operating environment factors, the least square method is used to correct the initially calculated propagation velocity to obtain the corrected ultrasonic propagation velocity.

[0034] Based on the corrected ultrasonic propagation velocity and the position information of the intelligent ultrasonic sensor, the triangulation positioning method is used to calculate the specific location of the particle defect. According to the time difference and propagation velocity of the signal reaching at least three sensors at different locations, the triangular area with the sensor as the vertex is determined. Then, the position coordinates of the defect in the GIL are obtained through geometric calculation, and the calculated defect position information is output.

[0035] A further improvement of the technical solution of the present invention is that the process of obtaining the position coordinates of the defect is:

[0036] Obtain the arrival time of ultrasonic signals from at least three intelligent ultrasonic sensors in the intelligent ultrasonic sensor network. The ultrasonic signal arrival time is the timestamp of the ultrasonic signal generated by the particle defect detected by the intelligent ultrasonic sensor. The time of the intelligent ultrasonic sensor that first detected the ultrasonic signal is determined as the reference time. This intelligent ultrasonic sensor is used as the reference sensor. The time difference between other sensors and the reference sensor is then calculated, and the time difference in the ultrasonic signal propagation from the defect location to different sensors is analyzed.

[0037] The corrected ultrasonic propagation velocity is used in combination with the time difference to calculate the signal propagation distance, which is the distance from the defect location to the second intelligent ultrasonic sensor and the distance from the defect location to the third intelligent ultrasonic sensor. The position coordinates of the three intelligent ultrasonic sensors are then determined, where the position coordinates of the reference sensor are known and fixed.

[0038] Based on the calculated distance of the ultrasonic signal from the defect location to the second intelligent ultrasonic sensor, the distance of the ultrasonic signal from the defect location to the third intelligent ultrasonic sensor, and the sensor coordinates, a geometric equation is established to solve the defect location, and a group of equations is obtained. By solving the above group of equations, the position coordinates of the defect are obtained, and then the calculated defect location information is output for use by the monitoring and early warning module.

[0039] A further improvement of the technical solution of the present invention is that the monitoring and early warning module specifically includes:

[0040] Receive data from the defect identification and positioning module in real time, including ultrasonic signal characteristic parameters of defects, defect types and defect locations, organize and visualize the data, and present it to operation and maintenance personnel through intuitive charts and graphical interfaces;

[0041] Based on the preset warning rules and thresholds, the system conducts risk analysis on the identified and located defects. When the ultrasonic signal characteristic parameters of the defect exceed the warning threshold, the warning mechanism is immediately triggered and the operation and maintenance personnel are notified through audible and visual alarms, SMS notifications, email reminders, etc.

[0042] Based on the ultrasonic signal characteristic parameters, defect type and defect location information of the defect, combined with the GIL operation and maintenance manual and historical maintenance data, targeted maintenance decision-making recommendations are generated, providing a basis for operation and maintenance personnel to formulate maintenance plans, help arrange maintenance work, and ensure the stable operation of the GIL.

[0043] In a second aspect, a method for online monitoring of GIL particle defects based on an intelligent sensing system is implemented based on an online monitoring system for GIL particle defects based on an intelligent sensing system, comprising the following steps:

[0044] S1. Deploy multiple intelligent ultrasonic sensors on the GIL to form a comprehensive sensor network to collect ultrasonic signals.

[0045] S2. Preprocess the collected ultrasonic signal, extract characteristic parameters that characterize the defect signal characteristics, and determine the baseline value of each characteristic parameter;

[0046] S3. Build a defect recognition model using machine learning algorithms, identify particle defects based on characteristic parameters, and locate defects using intelligent ultrasonic sensor position information and corrected propagation velocity.

[0047] S4. Analyze the identified particle defects according to the preset warning rules. When the characteristic parameters exceed the preset warning threshold, the warning mechanism is triggered, defect information is issued, and the operation and maintenance personnel are notified;

[0048] S5. Based on defect information, combined with the GIL operation and maintenance manual and historical maintenance data, generate maintenance decision recommendations to guide operation and maintenance personnel to formulate and implement maintenance plans.

[0049] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art:

[0050] 1. The present invention provides an online monitoring system and method for GIL particle defects based on an intelligent sensing system. By deploying multiple intelligent ultrasonic sensors on the GIL to form a sensor network, an intelligent algorithm is used to correct the ultrasonic propagation velocity error, and a triangulation positioning method is used to accurately calculate the defect location. The system and method comprehensively consider the sensor position information and signal propagation characteristics, effectively overcoming the positioning inaccuracy caused by propagation velocity uncertainty and sensor installation deviation in traditional methods, and significantly improving the accuracy of defect positioning.

[0051] 2. The present invention provides an online monitoring system and method for GIL particle defects based on an intelligent sensing system. By extracting a variety of characteristic parameters that characterize the characteristics of defect signals from preprocessed ultrasonic signals and constructing a defect recognition model, the characteristic parameters are analyzed through a machine learning algorithm, and the characteristic patterns of different types of particle defects are learned, thereby achieving accurate classification and identification of defects. This not only improves the accuracy of defect identification, but also enhances the system's ability to identify complex defect types, which helps to promptly discover and deal with potential safety hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0053] Figure 1 This is a schematic diagram of the functional modules of the system of the present invention;

[0054] Figure 2 Schematic diagram of the workflow of the present invention. DETAILED DESCRIPTION

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] Example 1, as Figure 1 、 Figure 2 As shown, the present invention provides a GIL particle defect online monitoring system based on an intelligent sensing system, including a defect monitoring platform, which is communicatively connected to an intelligent sensor network module, a signal conditioning module, a feature extraction module, a defect identification and positioning module, and a monitoring and early warning module, wherein the modules are electrically connected;

[0057] The intelligent sensor network module is used to deploy multiple intelligent ultrasonic sensors on the GIL to form a sensor network to collect the original ultrasonic signals inside the GIL, cover the monitoring area, reduce the impact of signal attenuation, and improve the comprehensiveness of signal acquisition. According to the structural layout, length, insulator distribution and operating environment factors of the GIL, the deployment location of the intelligent ultrasonic sensor is determined by combining computer simulation with on-site investigation to ensure that the intelligent ultrasonic sensor can cover the entire monitoring area. At the same time, the signal attenuation caused by excessive distance affects the detection effect. For the connection points, bends and other parts of the GIL that are prone to defects, the sensor density is appropriately increased. The intelligent ultrasonic sensor is fixed on the GIL shell according to the planned position. During installation, ensure that the sensor is in good contact with the GIL shell to reduce signal propagation. Interference and attenuation during the process. After installation, each intelligent ultrasonic sensor is individually debugged, including checking key parameters such as sensor sensitivity, frequency response range and directionality. At the same time, the receiving direction and angle of the intelligent ultrasonic sensor are calibrated so that it can accurately capture the ultrasonic signal inside the GIL. Each intelligent ultrasonic sensor is connected wirelessly to form an intelligent sensor network. All installed intelligent ultrasonic sensors are connected to the communication network of the monitoring system, and each intelligent ultrasonic sensor is used to collect the original ultrasonic signal inside the GIL in real time to achieve collaborative work between sensors. In view of the fact that the particle defects inside the GIL are distributed along the axial and radial directions of the pipeline, the GIL is regarded as a quasi-two-dimensional structure, and the sensors are arranged along the axial and radial directions of the GIL.

[0058] The signal conditioning module is used to perform pre-processing operations including filtering and amplification on the collected original ultrasonic signal, suppress noise, enhance effective signals, improve signal quality, reduce noise interference, and improve signal recognizability. The collected original ultrasonic signal is preliminarily filtered to remove high-frequency interference and low-frequency noise. The noise components beyond the frequency range of the ultrasonic signal are filtered out by hardware filters. At the same time, the analog filtering circuit is used to further weaken random noise and background interference, so that the ultrasonic signal is initially focused within the target frequency range. The ultrasonic signal after preliminary filtering is amplified. According to the initial strength of the ultrasonic signal, the corresponding amplification factor is selected to increase the signal amplitude to a level suitable for subsequent processing. The amplified ultrasonic signal is linearly compensated to correct nonlinear distortion during the amplification process and maintain the integrity of the ultrasonic signal. The amplified signal is finely filtered to further remove residual noise components. The adaptive filter dynamically adjusts the filter parameters according to the actual characteristics of the signal to suppress residual noise interference. At the same time, the filtered signal is smoothed to remove existing glitches and abnormal fluctuations, ultimately optimizing the signal quality.

[0059] The feature extraction module is used to extract characteristic parameters characterizing the characteristics of the defect signal from the preprocessed ultrasonic signal, determine the baseline value of each characteristic parameter, form a defect feature sequence, and select characteristic parameters for characterizing the characteristics of the particle defect from the preprocessed ultrasonic signal. Among them, according to the characteristics of the GIL particle defect signal, characteristic parameters including signal amplitude, signal frequency, signal duration, number of discharges, signal rise time and entropy of the signal time-frequency distribution are selected. The signal amplitude is the maximum voltage value of the ultrasonic signal on the time axis, reflecting the strength of the signal. By measuring the voltage waveform of the ultrasonic signal, its peak voltage value is directly read. The signal frequency is the number of times the signal completes periodic changes per unit time, describing the vibration speed of the signal. By performing spectral analysis on the ultrasonic signal, the main frequency components of the signal and their corresponding frequency values are obtained. The signal duration is the length of time the ultrasonic signal experiences from the beginning to the end, which is represented by measuring the time interval from the starting point to the end point of the signal. The number of discharges is the number of local discharges caused by particle defects per unit time, which is obtained by counting the discharge pulses within a period of time. The signal rise time is the time from the amplitude of the ultrasonic signal to the end point of the signal. The time required for the amplitude to rise from 10% to 90% reflects the speed of the signal rise. By measuring the time interval from the 10% point to the 90% point of the ultrasonic signal, the entropy of the signal time-frequency distribution is a quantification of the concentration of the signal time-frequency distribution, reflecting the complexity of the signal in the time-frequency domain. The smaller the entropy value, the more concentrated the signal energy is, which is related to a single defect source. The larger the entropy value, the more dispersed the signal energy distribution is, which is related to multiple defect sources. The pre-processed ultrasonic signal is analyzed to extract the selected characteristic parameters. Among them, the signal amplitude, signal duration and signal rise time are obtained through time domain analysis. , use frequency domain analysis to determine the signal frequency, count the number of discharges by pulse counting, and use wavelet transform tools to calculate the entropy of the signal time-frequency distribution. During the extraction process, ensure the accuracy and stability of each characteristic parameter to avoid extraction errors caused by signal fluctuations or noise interference. According to the extracted characteristic parameters, determine the baseline value of each characteristic parameter, that is, the reference value under normal operating conditions. Through comparative analysis, arrange the extraction results of each characteristic parameter in chronological order to form a defect feature sequence. Among them, the setting of the baseline value is based on the ultrasonic signal characteristics of the GIL during normal operation, which is obtained through long-term monitoring data statistics;

[0060] Defect identification and positioning module, which is used to combine defect feature sequences and pre-built defect identification models to identify particle defects and locate the locations of defects to improve positioning accuracy;

[0061] The monitoring and early warning module is used to display monitoring data, analyze results, issue early warnings for identified and located defects, and provide maintenance decision-making recommendations, thereby realizing intelligent monitoring and management of GIL particle defects and ensuring the safe and stable operation of the power system.

[0062] Example 2, as Figure 1 、 Figure 2 As shown, based on embodiment 1, the present invention provides a technical solution: preferably, the defect recognition and positioning module includes a defect pattern recognition unit and a defect positioning unit;

[0063] The defect pattern recognition unit is used to build a defect recognition model using a machine learning algorithm, analyze characteristic parameters, identify particle defects, collect a large number of known types of GIL particle defect samples, covering normal conditions and defective conditions, extract their characteristic parameters and label the defect types, build a comprehensive data set, and divide it into a training set and a test set. A machine learning algorithm based on a support vector machine model is selected as the basic architecture for building the defect recognition model. The support vector machine model is trained using the training set. By adjusting the algorithm parameters, the model performance is optimized to improve the accuracy and robustness of classification and recognition. The characteristic patterns of different types of particle defects are learned. The trained model is tested using the test set and the model's classification accuracy, recall rate, and F1 score are evaluated. Based on the evaluation results, the advantages and disadvantages of the model are analyzed and the model is further optimized to improve the model's generalization ability and defect recognition ability. The final defect recognition model is then applied to actual GIL particle defect monitoring. The extracted characteristic parameters are classified and recognized in real time to determine whether there is a defect and determine the defect type that matches the characteristic parameters.

[0064] The defect location unit is used to correct the propagation velocity error based on the identified particle defects and combine the position information and characteristic parameters of the intelligent ultrasonic sensors to locate the specific location of the defects. After identifying the particle defects, the defect location unit extracts the position information provided by the intelligent ultrasonic sensors and determines the specific coordinates of each intelligent ultrasonic sensor in the GIL. At the same time, based on the theoretical propagation velocity formula of ultrasonic waves in the GIL medium and combined with the initial conditions of the equipment material and ambient temperature, it preliminarily calculates the propagation velocity of ultrasonic waves in the GIL.

[0065] The propagation speed formula is:

[0066]

[0067] Where v is the propagation velocity of ultrasound in the GIL medium, E is the elastic modulus of the GIL medium, and ρ is the density of the GIL medium. As E increases, v increases; as ρ increases, v decreases.

[0068] According to the propagation characteristics of ultrasonic signals in GIL and the actual operating environment factors, the least square method is used to correct the initially calculated propagation velocity to obtain the corrected ultrasonic propagation velocity.

[0069] The corrected expression of ultrasonic propagation velocity is:

[0070] v corrected =v initial +Δv;

[0071]

[0072] Where, v corrected is the corrected ultrasonic propagation velocity, v initial is the initially calculated ultrasonic propagation velocity, Δv is the correction value, which is calculated by the least square method, and t i is the actual measured signal arrival time, According to the initial calculation speed v initial Calculated signal arrival time, d i is the distance the signal travels, n is the number of sensors involved in the correction calculation, and the corrected speed v corrected It should be closer to the actual propagation speed, with a deviation of less than 1%. Based on the corrected ultrasonic propagation speed and the position information of the intelligent ultrasonic sensor, the triangulation method is used to calculate the specific location of the particle defect. According to the time difference and propagation speed of the signal reaching at least three sensors at different locations, the triangular area with the sensor as the vertex is determined. Then, the position coordinates of the defect in the GIL are obtained through geometric calculation, and the calculated defect position information is output;

[0073] In addition, the process of obtaining the position coordinates of the defect is:

[0074] Obtain the arrival time of ultrasonic signals of at least three intelligent ultrasonic sensors in the intelligent ultrasonic sensor network. The arrival time of ultrasonic signals is the timestamp of ultrasonic signals generated by the detection of particle defects by the intelligent ultrasonic sensor, and determine the time of the intelligent ultrasonic sensor that detects the ultrasonic signal earliest as the reference time. This intelligent ultrasonic sensor is the reference sensor, and then calculate the time difference between other sensors and the reference sensor, analyze the time difference of ultrasonic signals propagating from the defect position to different sensors, use the corrected ultrasonic propagation speed, and combine the time difference to calculate the signal propagation distance, which are the distance of the ultrasonic signal from the defect position to the second intelligent ultrasonic sensor and the distance of the ultrasonic signal from the defect position to the third intelligent ultrasonic sensor. The position coordinates of the three intelligent ultrasonic sensors are determined, wherein the position coordinates of the reference sensor are known and fixed. According to the calculated distance of the ultrasonic signal from the defect position to the second intelligent ultrasonic sensor, the distance of the ultrasonic signal from the defect position to the third intelligent ultrasonic sensor, and the sensor coordinates, a geometric equation is established to solve the defect position, and a set of equations is obtained. By solving the above set of equations, the position coordinates of the defect are obtained, and then the calculated defect position information is output for use by the monitoring and early warning module. In addition, since the overall data volume of the three-dimensional space is too large, it is simplified to a plane for positioning. By measuring the time difference of the signal reaching each intelligent ultrasonic sensor, the position coordinates of the defect on the plane are calculated using the triangulation positioning method;

[0075] The expression of the equation system is:

[0076]

[0077] Where (x, y) is the position coordinate of the defect, v corrected is the corrected ultrasonic propagation speed, t1, t2, and t3 are the times when the ultrasonic signal reaches the intelligent ultrasonic sensor, and the coordinates of the three intelligent ultrasonic sensors are (x1, y1), (x2, y2), and (x3, y3), respectively;

[0078] The monitoring and early warning module specifically includes:

[0079] It receives data from the defect identification and positioning module in real time, including the ultrasonic signal characteristic parameters of the defect, the defect type and the defect location information, and organizes and visualizes the data, and displays it to the operation and maintenance personnel through intuitive charts and graphical interfaces. It conducts risk analysis on the identified and located defects according to the preset warning rules and warning thresholds. When the ultrasonic signal characteristic parameters of the defect exceed the warning threshold, the warning mechanism is immediately triggered, and the operation and maintenance personnel are informed through sound and light alarms, SMS notifications, email reminders, etc. Based on the ultrasonic signal characteristic parameters, defect type and defect location information of the defect, combined with the GIL operation and maintenance manual and historical maintenance data, it generates targeted maintenance decision suggestions, provides a basis for the operation and maintenance personnel to formulate maintenance plans, helps arrange maintenance work, and ensures the stable operation of the GIL.

[0080] Example 3, as Figure 1 、 Figure 2 As shown, based on Examples 1-2, the present invention further provides a method for online monitoring of GIL particle defects based on an intelligent sensing system, which is implemented based on an online monitoring system for GIL particle defects based on an intelligent sensing system, and includes the following steps:

[0081] S1. Deploy multiple intelligent ultrasonic sensors on the GIL to form a comprehensive sensor network to collect ultrasonic signals.

[0082] S2. Preprocess the collected ultrasonic signal, extract characteristic parameters that characterize the defect signal characteristics, and determine the baseline value of each characteristic parameter;

[0083] S3. Build a defect recognition model using machine learning algorithms, identify particle defects based on characteristic parameters, and locate defects using intelligent ultrasonic sensor position information and corrected propagation velocity.

[0084] S4. Analyze the identified particle defects according to the preset warning rules. When the characteristic parameters exceed the preset warning threshold, the warning mechanism is triggered, defect information is issued, and the operation and maintenance personnel are notified;

[0085] S5. Based on defect information, combined with the GIL operation and maintenance manual and historical maintenance data, generate maintenance decision recommendations to guide operation and maintenance personnel to formulate and implement maintenance plans.

[0086] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. An online monitoring system for GIL particle defects based on an intelligent sensing system, including a defect monitoring platform, characterized by: The defect monitoring platform is communicatively connected to an intelligent sensor network module, a signal conditioning module, a feature extraction module, a defect identification and positioning module, and a monitoring and early warning module, wherein electrical signals are connected between the modules; The intelligent sensor network module is used to deploy multiple intelligent ultrasonic sensors on the GIL to form a sensor network to collect the original ultrasonic signals inside the GIL; The signal conditioning module is used to perform pre-processing operations on the collected original ultrasonic signals; The feature extraction module is used to extract characteristic parameters representing the characteristics of the defect signal from the preprocessed ultrasonic signal, and determine the baseline value of each characteristic parameter to form a defect feature sequence; The defect recognition and positioning module is used to identify particle defects and locate the positions of the defects by combining the defect feature sequence and the pre-built defect recognition model; The monitoring and early warning module is used to display monitoring data and analysis results, issue early warnings for identified and located defects, and provide maintenance decision suggestions.

2. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 1, characterized in that: The intelligent sensor network module specifically includes: Determine the deployment location of intelligent ultrasonic sensors based on the GIL's structural layout, length, insulator distribution, and operating environment through a combination of computer simulation and on-site investigation. Fix the smart ultrasonic sensors on the GIL housing according to the planned locations. After installation, debug each smart ultrasonic sensor individually. Calibrate the receiving direction and angle of each smart ultrasonic sensor. Connect each smart ultrasonic sensor wirelessly to form a smart sensor network. All installed intelligent ultrasonic sensors are connected to the communication network of the monitoring system, and each intelligent ultrasonic sensor is used to collect the original ultrasonic signals inside the GIL in real time.

3. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 1, characterized in that: The signal conditioning module specifically includes: Perform preliminary filtering on the collected raw ultrasonic signals, and use analog filtering circuits to further reduce random noise and background interference; Amplify the ultrasonic signal after preliminary filtering, select the corresponding amplification factor according to the initial intensity of the ultrasonic signal, and perform linear compensation on the amplified ultrasonic signal; The amplified signal is finely filtered to further remove the residual noise components, and the filtered signal is smoothed to remove the existing burrs and abnormal fluctuations.

4. The GIL particle defect online monitoring system based on an intelligent sensing system according to claim 1, characterized in that: The feature extraction module specifically includes: Selecting characteristic parameters for characterizing particle defect characteristics from the preprocessed ultrasonic signal, wherein the characteristic parameters selected include signal amplitude, signal frequency, signal duration, number of discharges, signal rise time, and entropy of the signal time-frequency distribution based on the characteristics of the GIL particle defect signal; The pre-processed ultrasonic signal is analyzed to extract selected characteristic parameters. The signal amplitude, signal duration and signal rise time are obtained through time domain analysis. The signal frequency is determined by frequency domain analysis. The number of discharges is counted by pulse counting. The entropy of the signal time-frequency distribution is calculated using wavelet transform tools. Based on the extracted characteristic parameters, the baseline value of each characteristic parameter is determined, that is, the reference value under normal operating conditions. Through comparative analysis, the extraction results of each characteristic parameter are arranged in chronological order to form a defect characteristic sequence.

5. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 4, characterized in that: The defect recognition and positioning module includes a defect pattern recognition unit and a defect positioning unit; The defect pattern recognition unit is used to construct a defect recognition model using a machine learning algorithm, analyze characteristic parameters, and identify particle defects; The defect locating unit is used to correct the propagation velocity error and locate the specific position of the defect based on the identified particle defect and in combination with the position information and characteristic parameters of the intelligent ultrasonic sensor.

6. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 5, characterized in that: The defect pattern recognition unit specifically includes: Collect a large number of GIL particle defect samples of known types, covering both normal and defective conditions, extract their characteristic parameters and label the defect types, construct a comprehensive dataset, and divide it into training and test sets; A machine learning algorithm based on a support vector machine model was selected as the basic architecture for building the defect recognition model. The support vector machine model was trained using a training set to learn the characteristic patterns of different types of particle defects. The trained model was tested using a test set to evaluate the model's classification accuracy, recall rate, and F1 score. Based on the evaluation results, the model was further optimized to obtain the final defect recognition model. The defect recognition model is applied to actual GIL particle defect monitoring, and the extracted characteristic parameters are classified and identified in real time to determine whether there are defects and determine the defect type that matches the characteristic parameters.

7. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 5, characterized in that: The defect location unit specifically includes: After identifying the particle defects, the defect location unit extracts the position information provided by the intelligent ultrasonic sensors and determines the specific coordinates of each intelligent ultrasonic sensor in the GIL. At the same time, based on the theoretical propagation speed formula of ultrasonic waves in the GIL medium, combined with the initial conditions of the equipment material and ambient temperature, it preliminarily calculates the propagation speed of ultrasonic waves in the GIL. According to the propagation characteristics of ultrasonic signals in GIL and the actual operating environment factors, the least square method is used to correct the initially calculated propagation velocity to obtain the corrected ultrasonic propagation velocity. Based on the corrected ultrasonic propagation velocity and the position information of the intelligent ultrasonic sensor, the triangulation positioning method is used to calculate the specific location of the particle defect. According to the time difference and propagation velocity of the signal reaching at least three sensors at different locations, the triangular area with the sensor as the vertex is determined. Then, the position coordinates of the defect in the GIL are obtained through geometric calculation, and the calculated defect position information is output.

8. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 7, characterized in that: The process of obtaining the position coordinates of the defect is as follows: Obtain the arrival time of ultrasonic signals from at least three intelligent ultrasonic sensors in the intelligent ultrasonic sensor network. The ultrasonic signal arrival time is the timestamp of the ultrasonic signal generated by the particle defect detected by the intelligent ultrasonic sensor. The time of the intelligent ultrasonic sensor that first detected the ultrasonic signal is determined as the reference time. This intelligent ultrasonic sensor is used as the reference sensor. The time difference between other sensors and the reference sensor is then calculated, and the time difference in the ultrasonic signal propagation from the defect location to different sensors is analyzed. The corrected ultrasonic propagation velocity is used in combination with the time difference to calculate the signal propagation distance, which is the distance from the defect location to the second intelligent ultrasonic sensor and the distance from the defect location to the third intelligent ultrasonic sensor. The position coordinates of the three intelligent ultrasonic sensors are then determined, where the position coordinates of the reference sensor are known and fixed. Based on the calculated distance of the ultrasonic signal from the defect location to the second intelligent ultrasonic sensor, the distance of the ultrasonic signal from the defect location to the third intelligent ultrasonic sensor, and the sensor coordinates, a geometric equation is established to solve the defect location, and a group of equations is obtained. By solving the above group of equations, the position coordinates of the defect are obtained, and then the calculated defect location information is output for use by the monitoring and early warning module.

9. The GIL particle defect online monitoring system based on the intelligent sensing system according to claim 7, characterized in that: The monitoring and early warning module specifically includes: Receive data from the defect identification and positioning module in real time, including ultrasonic signal characteristic parameters of defects, defect types and defect locations, organize and visualize the data, and present it to operation and maintenance personnel through intuitive charts and graphical interfaces; Based on the preset warning rules and warning thresholds, the system conducts risk analysis on the identified and located defects. When the ultrasonic signal characteristic parameters of the defect exceed the warning threshold, the warning mechanism is immediately triggered to inform the operation and maintenance personnel. Based on the ultrasonic signal characteristic parameters, defect type and defect location information of the defect, combined with the GIL operation and maintenance manual and historical maintenance data, targeted maintenance decision recommendations are generated, providing a basis for operation and maintenance personnel to formulate maintenance plans.

10. A method for online monitoring of GIL particle defects based on an intelligent sensor system, implemented based on the online monitoring system for GIL particle defects based on an intelligent sensor system according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1. Deploy multiple intelligent ultrasonic sensors on the GIL to form a comprehensive sensor network to collect ultrasonic signals. S2. Preprocess the collected ultrasonic signal, extract characteristic parameters that characterize the defect signal characteristics, and determine the baseline value of each characteristic parameter; S3. Build a defect recognition model using machine learning algorithms, identify particle defects based on characteristic parameters, and locate defects using intelligent ultrasonic sensor position information and corrected propagation velocity. S4. Analyze the identified particle defects according to the preset warning rules. When the characteristic parameters exceed the preset warning threshold, the warning mechanism is triggered, defect information is issued, and the operation and maintenance personnel are notified; S5. Based on defect information, combined with the GIL operation and maintenance manual and historical maintenance data, generate maintenance decision recommendations to guide operation and maintenance personnel to formulate and implement maintenance plans.

Images