Substation cable insulation monitoring method for realizing early warning

Abstract

The invention discloses a transformer substation cable insulation monitoring method for realizing early warning. The method comprises the steps of deployment of a sensor network, sound signal analysis, abnormal sound analysis and evaluation, insulation hidden danger positioning, early warning, emergency and the like. According to the invention, through combination of various sensors and signal processing technologies, comprehensive monitoring of the insulation state of the transformer substation cable is realized, the accuracy and reliability of monitoring are improved, early signs of cable insulation problems can be captured in real time, sufficient time is provided for operation and maintenance personnel to carry out intervention and maintenance, and the operation and maintenance efficiency is improved. Therefore, potential electric power accidents are effectively avoided, safe and stable operation of a power grid is ensured, abnormal sound signals in cable operation can be accurately recognized through the sound sensor in the aspects of sound signal collection and analysis, deep analysis is conducted on the abnormal sound signals, and the accuracy of insulation problem detection is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of substation cable insulation monitoring, and particularly to a substation cable insulation monitoring method for realizing early warning. Background Art

[0002] With the continuous development of the power industry and the continuous expansion of the power grid scale, the substation, as a crucial part of the power system, undertakes the important tasks of power transmission, distribution, and conversion. In the substation, the cable, as the main medium for connecting various devices and transmitting electric energy, its insulation state is directly related to the safe and stable operation of the power system. However, due to the long-term operation of the cable in a complex and changeable electromagnetic environment and being often affected by environmental factors such as temperature and humidity, the insulation performance of the cable gradually deteriorates, and even insulation faults are caused, posing a serious threat to the safe and stable operation of the power system. Therefore, real-time monitoring and early warning of the insulation state of substation cables, and timely discovery and handling of potential insulation hidden dangers are of great significance for ensuring the safe and stable operation of the power system.

[0003] Traditional substation cable insulation monitoring methods mainly rely on regular power outage detection and manual inspection, which are not only time-consuming and laborious, but also unable to monitor the insulation state of the cable in real time, making it difficult to discover and handle potential insulation hidden dangers in a timely manner. In addition, due to the complexity of the cable structure and the particularity of the operating environment, traditional monitoring methods often have difficulty accurately judging the location and degree of insulation faults, bringing great difficulties to subsequent maintenance and handling work. Therefore, traditional substation cable insulation monitoring methods can no longer meet the high requirements of modern power systems for safety and stability. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a substation cable insulation monitoring method for realizing early warning. This method realizes precise evaluation and hidden danger location of the cable insulation state through comprehensive and multi-parameter real-time monitoring of substation cables, using voiceprint recognition, time series analysis, and signal propagation theory technologies. Compared with traditional technologies, this method can discover insulation problems earlier, improve the accuracy of early warning, and reduce the failure rate.

[0005] To solve the above technical problems, the present invention provides a substation cable insulation monitoring method for realizing early warning, which includes the following steps:

[0006] Step S10, according to the substation layout, electromagnetic environment, and cable characteristics, install sound, partial discharge, temperature, and electromagnetic interference sensors at key positions of the cable, construct a wireless communication sensor network, and transmit the collected data to the monitoring center in real time;

[0007] Step S11: Set the sensor sampling parameters, continuously collect sound signals, perform filtering, amplification, and digitization processing, extract frequency features, calculate the energy distribution features, compare with the normal operation sound feature template, and mark the suspected abnormal signals;

[0008] Step S12: Establish a time series model, analyze the change trend of the characteristic parameters of the suspected abnormal sound signals, predict the development direction, and conduct correlation analysis on the abnormal characteristics of the sound signals and the temperature, electromagnetic interference, and other sensor data collected simultaneously to comprehensively evaluate the cable insulation problem;

[0009] Step S13: Integrate the data of multiple sensors, establish a mathematical model of signal propagation, consider the cable physical structure, medium characteristics, and environmental factors, compensate and correct the signal propagation speed, and calculate the coordinate of the insulation hidden danger location;

[0010] Step S14: Dynamically set the warning threshold, adjust it in real time according to the statistical analysis results of the monitoring data, generate a warning message when the insulation problem exceeds the threshold, publish it to the operation and maintenance personnel and management personnel through multiple channels, and automatically start the corresponding emergency response plan.

[0011] Preferably, the step S11 further includes:

[0012] Set the sampling parameters of the sensor according to different cable types and operating environments, start the sensor to continuously collect sound signals, perform filtering, amplification, and digitization processing on the collected sound signals, extract frequency features using time-frequency analysis methods, and calculate the energy distribution features based on statistical methods. Compare these features with the normal operation sound feature template of the cable, and mark the suspected abnormal signals corresponding to the characteristic values that significantly deviate from the normal range.

[0013] Preferably, in the step S11, it includes:

[0014] The signal energy is calculated in the following manner:

[0015] Assume the collected sound signal is S(t), and its discrete form is S[n]. N is the number of sampling points. Calculate the energy E of the signal. The formula is: Assume the normal operation energy mean is E avg and the energy standard deviation is σ E Calculate the energy anomaly coefficient C E The formula is: where ∈ is an extremely small positive number to prevent the denominator from being zero;

[0016] The frequency distribution characteristics are calculated in the following manner:

[0017] Perform a fast Fourier transform on the sound signal to obtain the spectrum F[k], and calculate the centroid frequency f of the spectrum center, the formula is: Define the average value of the center of gravity frequency during normal operation as f avg , and the standard deviation as σ f , calculate the frequency anomaly coefficient C f , the formula is:

[0018]

[0019] Judge whether the signal is abnormal in the following way:

[0020] Define the abnormality degree A of the sound signal S , comprehensively consider the energy and frequency anomaly coefficient to calculate the abnormality degree of the sound signal, and the formula is: When A S exceeds the preset normal threshold T AS , it is judged that the sound signal is abnormal.

[0021] Preferably, in the step S12, it includes:

[0022] Measure the influence degree of environmental factors on the sound signal through the correlation formula between the sound signal and environmental factors. Let the environmental temperature be T, the environmental humidity be H, the electromagnetic interference intensity be I, and the amplitude of the k-th frequency component of the sound signal S(t) in the frequency domain be A S,k , define a correlation coefficient vector C = [C T , C H , C I , where: The comprehensive correlation degree R between the sound signal and environmental factors env is:

[0023] Preferably, in the step S12, it includes:

[0024] Predict the probability of the current cable having insulation problems by using historical data through the insulation problem probability prediction formula. Let the historical data record N d detection situations, among which there are n prob times of insulation problems. For a set of feature vectors X = [x1, x2,..., x m detected currently, the feature vector corresponding to each historical data is X i = [x i1 , x i2 ,..., x im , and define the feature distance function as: For a distance threshold d th , find the subset of historical data within the range of d th from the current feature vector X, and its quantity is n near , among which there are n near-probIf there is an insulation problem for the first time, the predicted probability P of the current insulation problem prob is as follows:

[0025] Preferably, the step S13 further includes:

[0026] Considering the influence of the actual physical structure of the cable, dielectric characteristics, and surrounding environmental factors on signal propagation, a signal propagation mathematical model is established by fusing multiple sensor data. The influence of the number of cable bends, the number of joints, and the number of surrounding obstacles on the complexity of the signal propagation path is evaluated, the signal propagation speed is compensated and corrected, the signal propagation model is optimized, and by defining a positioning objective function, the minimum value of this objective function is solved using an optimization algorithm, so as to obtain the position coordinates of the insulation hidden danger.

[0027] Preferably, in the step S13, it includes:

[0028] When establishing the signal propagation mathematical model, assume that there are M sensors in total, and the signal amplitude collected by the i-th sensor is A i , and the phase is The distance from the signal source to the i-th sensor is d i , and the signal propagation speed is v;

[0029] Calculate the signal propagation delay time t i , and the formula is:

[0030] Calculate the signal propagation delay time. According to the signal propagation theory, the relationship between the signal amplitude and the propagation distance is expressed as: where A0 is the signal source amplitude, α is a coefficient related to the cable attenuation characteristic, and the relationship between the phase and the distance is: where ω is the signal angular frequency, is the initial phase of the signal source;

[0031] Construct the positioning objective function J as:

[0032]

[0033] Use the gradient descent method to solve the minimum value of the objective function. Assume that the initial estimated value of the insulation hidden danger position coordinates is x 0 , set the learning rate η, the iteration number threshold T, and the convergence accuracy ∈, and take the partial derivative of the objective function J with respect to the coordinate variable x k to obtain the gradient vector as:

[0034] In each iteration n = 0, 1, 2,..., update the estimated value of the insulation hidden danger position coordinates according to the update formula of the gradient descent method, and the formula is:

[0035] After each iteration, compare the objective function value J(x n ) of the current iteration with the objective function value J(x n-1 ) of the previous iteration. If |J(x n ) - J(x n-1 )| < ∈, it is considered that the algorithm converges. When the algorithm converges, obtain the current estimated coordinate value of the insulation hidden danger location as the required coordinate of the insulation hidden danger location.

[0036] Preferably, in the step S13, it includes:

[0037] Compensate and correct the signal propagation speed by evaluating the influence of the number of cable bends, the number of joints, and the number of surrounding obstacles on the complexity of the signal propagation path:

[0038] Let the number of cable bends be B, the number of joints be J, the number of surrounding obstacles be O, and the signal propagation path complexity coefficient C path The calculation formula is: C path = α1B + α2J + α3O,

[0039] where α1, α2, and α3 are the corresponding influence weights. According to the value of C parh , make a more targeted correction to the signal propagation. When C path is relatively large, add a compensation term when calculating the signal propagation speed v: v = v0(1 - βC path ) where v0 is the theoretical speed without the influence of complex paths, and β is the compensation coefficient.

[0040] Preferably, in the step S14, it includes:

[0041] Dynamically adjust the warning threshold according to the warning threshold dynamic adjustment formula. Let the cable operation time be t run , the initial warning threshold be T init , the aging coefficient be β, and the dynamic warning threshold T dynamic The calculation formula is: T dynamic = T init ·(1 + β·t run ).

[0042] Preferably, in the step S14, it includes:

[0043] Dynamically adjust the warning threshold in combination with the real-time weather conditions. Let the real-time weather condition parameter vector be W = [w1, w2, …, w q , the weather influence coefficient vector be k w = [k w1 , k w2 , …, k wq , and the warning threshold is adjusted to:

[0044] Implementing the embodiments of the present invention has the following beneficial effects:

[0045] The present invention provides a method for monitoring the insulation of substation cables. By combining multiple sensors and signal processing technologies, it realizes the comprehensive monitoring of the insulation status of substation cables. This method not only improves the accuracy and reliability of monitoring, but also can capture the early signs of cable insulation problems in real time, providing sufficient time for operation and maintenance personnel to intervene and maintain, thus effectively avoiding potential power accidents and ensuring the safe and stable operation of the power grid.

[0046] In terms of the acquisition and analysis of sound signals, the present invention uses sound sensors to accurately identify abnormal sound signals during cable operation and conducts in-depth analysis on them, further improving the accuracy of insulation problem detection.

[0047] In addition, by establishing a mathematical model of signal propagation and fusing the data of multiple sensors, the present invention can accurately calculate the location of insulation hidden dangers. This not only improves the positioning accuracy, but also provides more accurate fault information for operation and maintenance personnel, helps to quickly locate and repair the fault point, reduces the power outage time and maintenance cost. At the same time, the present invention also considers the influence of the actual physical structure of the cable, dielectric characteristics and surrounding environmental factors on signal propagation, and improves the model through a combination of field experiments and numerical simulations, further enhancing the accuracy and reliability of positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0049] Figure 1 It is the main flowchart of an embodiment of a method for monitoring the insulation of substation cables to achieve early warning provided by the present invention;

[0050] Figure 2 For Figure 1 It is the flowchart for locating insulation hidden dangers in step S13 in DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0052] As shown Figure 1 in the figure, a main process schematic diagram of an embodiment of a substation cable insulation monitoring method for realizing early warning provided by the present invention is shown. In combination with Figure 2 shown in the figure, in this embodiment, the method includes the following steps:

[0053] Step S10, deployment of the sensor network: According to the substation layout, electromagnetic environment and cable characteristics, install sound, partial discharge, temperature and electromagnetic interference sensors at key positions of the cable, construct a wireless communication sensor network, and transmit the collected data to the monitoring center in real time;

[0054] Specifically, in the monitoring preparation stage, a detailed survey of the overall layout, electromagnetic environment, temperature and humidity change range of urban substations is carried out. Generally speaking, the electromagnetic environment is relatively complex, there are certain electromagnetic interference sources, the temperature change range is between -10°C and 40°C, the humidity change range is between 30% and 80%, and the main types of in-station cables are cross-linked polyethylene insulated cables and oil-paper insulated cables, and the voltage levels cover 10kV and 35kV; analyze the types, voltage levels, lengths, diameters, physical and electrical characteristics of the insulating materials of the in-station cables, and install sound sensors, partial discharge sensors, temperature sensors and electromagnetic interference sensors at key positions along the cable, at joints and at bends according to the analysis results, construct a sensor network through wireless communication technology, and transmit the collected data to the monitoring center in real time.

[0055] Step S11, sound signal analysis: Set the sensor sampling parameters, continuously collect sound signals, perform filtering, amplification and digital processing, extract frequency characteristics, calculate the energy distribution characteristics, compare with the normal operation sound characteristic template, and mark the suspected abnormal signals;

[0056] Specifically, the step S11 further includes:

[0057] Set the sampling parameters of the sensor according to different cable types and operating environments, start the sensor to continuously collect sound signals, perform filtering, amplification and digital processing on the collected sound signals, use time-frequency analysis methods to extract frequency characteristics, and calculate the energy distribution characteristics based on statistical methods, and compare these characteristics with the normal operation sound characteristic template of the cable, and mark the suspected abnormal signals corresponding to the characteristic values that significantly deviate from the normal range.

[0058] More specifically, in the step S11, it includes:

[0059] The signal energy is calculated in the following manner:

[0060] Let the collected sound signal be S(t), and its discrete form be S[n] (n = 0, 1, …, N - 1, N), where N is the number of sampling points. Calculate the energy E of the signal, and the formula is: Let the average energy during normal operation be E avg , and the standard deviation of energy be σ E , and calculate the energy anomaly coefficient C E , and the formula is: where ∈ is an extremely small positive number to prevent the denominator from being zero;

[0061] The following method is used to calculate the frequency distribution characteristics:

[0062] Perform a fast Fourier transform (FFT) on the sound signal to obtain the frequency spectrum Calculate the centroid frequency f of the frequency spectrum center , and the formula is: Define the average centroid frequency during normal operation as f avg , and the standard deviation as σ f , and calculate the frequency anomaly coefficient C f , and the formula is:

[0063] The following method is used to determine whether the signal is abnormal:

[0064] Define the abnormality degree A of the sound signal S , and comprehensively consider the energy and frequency anomaly coefficients to calculate the abnormality degree of the sound signal. The formula is: When A S exceeds the preset normal threshold T AS , it is determined that the sound signal is abnormal.

[0065] Step S12, abnormal sound analysis and evaluation: Establish a time series model, analyze the change trend of the characteristic parameters of the suspected abnormal sound signal, predict the development direction, and perform correlation analysis on the abnormal characteristics of the sound signal and the temperature, electromagnetic interference, and other sensor data collected simultaneously to comprehensively evaluate the cable insulation problem;

[0066] Among them, in the said step S12, it includes:

[0067] Use historical data to predict the probability of the current cable having an insulation problem through the insulation problem probability prediction formula. Suppose there are N d times of detection records in the historical data, among which there are n prob times of insulation problems. For a set of characteristic vectors X = [x1, x2, …, x m detected currently, the characteristic vector corresponding to each historical data is X i = [x i1 , x i2 , …, x im, define the feature distance function as: For a distance threshold d th , find the subset of historical data within the range of d from the current feature vector X, and the number of such data is n th , among which there are n near occurrences of insulation problems, then the predicted probability P near-prob of the current insulation problem is: prob That is:

[0068] In the step S12, it further includes:

[0069] Measure the influence degree of environmental factors on the sound signal through the correlation formula between the sound signal and environmental factors. Let the environmental temperature be T, the environmental humidity be H, the electromagnetic interference intensity be I, and the amplitude of the k-th frequency component of the sound signal S(t) in the frequency domain be A S,k , define a correlation coefficient vector C = [C T , C H , C I , where: The comprehensive correlation degree R env between the sound signal and environmental factors is: Based on this, analyze the correlation between the abnormal sound signal and other sensor data such as temperature and electromagnetic interference, and at the same time, combine the cable operation historical data to comprehensively evaluate the cable insulation problem.

[0070] Step S13, insulation hidden danger location: fuse the data of multiple sensors, establish a mathematical model of signal propagation, consider the cable physical structure, dielectric characteristics and environmental factors, compensate and correct the signal propagation speed, and calculate the position coordinates of the insulation hidden danger;

[0071] Among them, the step S13 further includes:

[0072] Consider the influence of the actual physical structure of the cable (such as cable bending, joint conditions), dielectric characteristics and surrounding environmental factors on signal propagation. Establish a mathematical model of signal propagation by fusing the data of multiple sensors, evaluate the influence of the number of cable bends, the number of joints and the number of surrounding obstacles on the complexity of the signal propagation path, compensate and correct the signal propagation speed, optimize the signal propagation model, and define a positioning objective function, and use an optimization algorithm to solve the minimum value of this objective function, so as to obtain the position coordinates of the insulation hidden danger.

[0073] More specifically, in the step S13, it includes:

[0074] Step S130, collect sensor data;

[0075] Step S131, when establishing the mathematical model of signal propagation, assume that there are M sensors in total, the signal amplitude collected by the i-th sensor is A i , and the phase is The distance from the signal source to the i-th sensor is d i , and the signal propagation speed is v;

[0076] According to the signal propagation theory, calculate the signal propagation delay time t i , and the formula is:

[0077] According to the signal propagation theory, the relationship between the signal amplitude and the propagation distance is expressed as: where A0 is the signal source amplitude, α is the coefficient related to the cable attenuation characteristic, and the relationship between the phase and the distance is: where ω is the signal angular frequency, is the initial phase of the signal source;

[0078] Step S132, by evaluating the influence of the number of cable bends, the number of joints, and the number of surrounding obstacles on the complexity of the signal propagation path, compensate and correct the signal propagation speed:

[0079] Assume that the number of cable bends is B, the number of joints is J, the number of surrounding obstacles is O, and the signal propagation path complexity coefficient C path The calculation formula of is: C path =α1B + α2J + α3O, where α1, α2, and α3 are the corresponding influence weights; according to C path value, make a more targeted correction to the signal propagation speed. When C path is relatively large, add a compensation term v = v0(1 - βC path ) when calculating the signal propagation speed v, where v0 is the theoretical speed without the influence of complex paths, β is the compensation coefficient, and calculate the coordinates of the insulation hidden danger location;

[0080] Step S133, construct the objective function J for positioning as:

[0081]

[0082] Step S134, use the gradient descent method to solve the minimum value of the objective function. Assume that the initial estimated value of the insulation hidden danger location coordinates is x 0 , set the learning rate η, the iteration number threshold T, and the convergence accuracy ∈, and take the partial derivative of the objective function J with respect to the coordinate variable x k to obtain the gradient vector as:

[0083] Step S135, in each iteration \(n = 0, 1, 2, \cdots\), update the estimated value of the insulation hidden danger position coordinates according to the update formula of the gradient descent method. The formula is:

[0084] After each iteration, compare the objective function value \(J(x\) n ) of the current iteration and the objective function value \(J(x\) n-1 ) of the previous iteration. If \(|J(x\) n ) - J(x\) n-1 )| < \(\epsilon\), it is considered that the algorithm converges. When the algorithm converges, obtain the current estimated value of the insulation hidden danger position coordinates as the required insulation hidden danger position coordinates;

[0085] Step S136, output the insulation hidden danger position coordinates.

[0086] Step S14, early warning and emergency response: Dynamically set the early warning threshold, adjust it in real time according to the statistical analysis results of the monitoring data, generate early warning information when the insulation problem exceeds the threshold, and send it to the operation and maintenance personnel and management personnel through multiple channels, and automatically start the corresponding emergency response plan.

[0087] Among them, in the said step S14, it includes:

[0088] Dynamically set different early warning thresholds according to factors such as the type of cable, operation years, and historical fault data. Let the operation time of the cable be \(t\) run , the initial early warning threshold is \(T\) init , the aging coefficient is \(\beta\), and the dynamic early warning threshold \(T\) dynamic The calculation formula is: \(T\) dynamic = \(T\) init ·(1 + \(\beta\cdot t\) run ).

[0089] In the said step S14, it includes:

[0090] Dynamically adjust the early warning threshold in combination with the real-time weather conditions. Let the real-time weather condition parameter vector be \(W = [w1, w2, \cdots, w\) q , and the weather influence coefficient vector be \(k\) w = \([k\) w1 , \(k\) w2 , \(\cdots\), \(k\) wq , and the early warning threshold is adjusted to: When the insulation problem exceeds the set threshold, detailed warning information is generated and the warning information is promptly released to the operation and maintenance personnel and relevant management personnel through multiple channels such as the in-station communication system, the SMS platform, and the mobile application. And according to the severity of the warning information, the corresponding emergency response plan is automatically activated. If it is judged to be a serious insulation problem, the corresponding emergency response plan is automatically activated, such as cutting off the power supply in the area where the faulty cable is located and dispatching repair personnel to carry the corresponding tools and equipment to the hidden danger location for maintenance.

[0091] It can be understood that, in summary: in the cable insulation monitoring of urban substations, through the deployment of a comprehensive sensor network, data is collected using a variety of sensors. In the stage of sound signal analysis, parameters such as signal energy and frequency characteristics are accurately calculated to determine whether the sound signal is abnormal. In the link of abnormal sound analysis and evaluation, historical data and environmental factors are used to deeply analyze the abnormal sound. When locating insulation hidden dangers, a suitable model is established and the minimum value is solved by the gradient descent method to determine the hidden danger location. At the same time, the propagation speed is corrected considering the path complexity. In terms of warning and emergency, the warning threshold is dynamically set and adjusted, combined with the weather conditions, to ensure the safe operation of urban substation cables, effectively identify abnormalities, locate hidden dangers, and dynamically warn, which can timely detect insulation problems and take corresponding measures, reduce the probability of faults, and ensure the stability of urban power supply.

[0092] Implementing the embodiments of the present invention has the following beneficial effects:

[0093] The present invention provides a method for monitoring the insulation of substation cables. By combining a variety of sensors and signal processing technologies, it realizes the comprehensive monitoring of the insulation state of substation cables. This method not only improves the accuracy and reliability of monitoring, but also can capture the early signs of cable insulation problems in real time, providing sufficient time for operation and maintenance personnel to intervene and maintain, thus effectively avoiding potential power accidents and ensuring the safe and stable operation of the power grid.

[0094] In terms of the acquisition and analysis of sound signals, the present invention uses sound sensors to accurately identify abnormal sound signals during cable operation and deeply analyze them, further improving the accuracy of insulation problem detection.

[0095] In addition, by establishing a mathematical model of signal propagation and integrating data from multiple sensors, the present invention can accurately calculate the location of insulation hidden dangers. This not only improves the positioning accuracy, but also provides more accurate fault information for operation and maintenance personnel, helping to quickly locate and repair the fault point, reducing the power outage time and maintenance cost. At the same time, the present invention also considers the influence of the actual physical structure of the cable, the medium characteristics, and the surrounding environmental factors on signal propagation, and improves the model through a combination of on-site experiments and numerical simulations, further enhancing the accuracy and reliability of positioning.

[0096] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a unit, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0097] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a unit for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0098] The above-disclosed is only a preferred embodiment of the present invention, and of course, it cannot be used to limit the scope of the rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A substation cable insulation monitoring method for realizing early warning, characterized in that, The method includes the following steps: Step S10: According to the substation layout, electromagnetic environment, and cable characteristics, install sound, partial discharge, temperature, and electromagnetic interference sensors at key positions of the cable, construct a wireless communication sensor network, and transmit the collected data to the monitoring center in real time; Step S11: Set the sensor sampling parameters, continuously collect sound signals, perform filtering, amplification, and digitization processing, extract frequency characteristics, calculate the energy distribution characteristics, compare with the normal operation sound characteristic template, and mark the suspected abnormal signals; Step S12: Establish a time series model, analyze the change trend of the characteristic parameters of the suspected abnormal sound signals, predict the development direction, and perform correlation analysis on the abnormal characteristics of the sound signals and the temperature, electromagnetic interference, and other sensor data collected simultaneously to comprehensively evaluate the cable insulation problem; Step S13: Integrate the data of multiple sensors, establish a signal propagation mathematical model, consider the cable physical structure, dielectric characteristics, and environmental factors, compensate and correct the signal propagation speed, and calculate the coordinate of the insulation hidden danger position; Step S14: Dynamically set the warning threshold, adjust it in real time according to the statistical analysis result of the monitoring data, generate a warning message when the insulation problem exceeds the threshold, publish it to the operation and maintenance personnel and management personnel through multiple channels, and automatically start the corresponding emergency response plan.

2. The method according to claim 1, wherein The step S11 further includes: Set the sampling parameters of the sensors for different cable types and operating environments, start the sensors to continuously collect sound signals, perform filtering, amplification, and digitization processing on the collected sound signals, extract frequency characteristics using time-frequency analysis methods, and calculate the energy distribution characteristics based on statistical methods. Compare these characteristics with the normal operation sound characteristic template of the cable, and mark the suspected abnormal signals corresponding to the characteristic values that deviate significantly from the normal range.

3. The method according to claim 2, characterized in that, In the step S11, it includes: The signal energy is calculated in the following manner: Let the collected sound signal be S(t), and its discrete form be S[n]. N is the number of sampling points. Calculate the energy E of the signal. The formula is as follows: Let the average value of the normal operating energy be E avg , and the standard deviation of the energy be σ E . Calculate the energy anomaly coefficient C E . The formula is as follows: where ∈ is an extremely small positive number to prevent the denominator from being zero; The frequency distribution characteristics are calculated in the following manner: The sound signal is subjected to a fast Fourier transform to obtain the spectrum F[k], and the center frequency f of the spectrum is calculated center , and the formula is: Define the average value of the center frequency during normal operation as f avg , and the standard deviation as σ f , calculate the frequency anomaly coefficient C f , and the formula is: The signal abnormality is judged in the following manner: Define the abnormality degree A of the sound signal S , comprehensively consider the energy and frequency abnormality coefficients to calculate the abnormality degree of the sound signal, and the formula is: When A S exceeds the preset normal threshold T AS , it is determined that the sound signal is abnormal.

4. The method according to claim 3, characterized in that, In the step S12, it includes: Measure the influence degree of environmental factors on the sound signal through the correlation formula between the sound signal and environmental factors. Let the environmental temperature be T, the environmental humidity be H, the electromagnetic interference intensity be I, and the amplitude of the k-th frequency component of the sound signal S(t) in the frequency domain be A S,k , define a correlation coefficient vector C = [C T , C H , C I , where: The comprehensive correlation degree R env between the sound signal and environmental factors is:

5. The method according to claim 4, characterized in that, In the step S12, it includes: Use historical data to predict the probability of insulation problems in the current cable through the insulation problem probability prediction formula. Assume that the historical data records N d detection cases, among which there are n prob cases with insulation problems. For a set of feature vectors X = [x1, x2, …, x m detected currently, the feature vector corresponding to each historical data is X i = [x i1 , x i2 , …, x im . Define the feature distance function as: For a distance threshold d th , find the subset of historical data within a distance d th from the current feature vector X. The number of such data is n near , among which there are n near-prob cases with insulation problems. Then the predicted probability P prob of the current insulation problem is:

6. The method according to claim 5, wherein The step S13 further includes: Consider the influence of the actual physical structure, dielectric characteristics, and surrounding environmental factors of the cable on signal propagation. Establish a signal propagation mathematical model by integrating the data of multiple sensors, evaluate the influence of the number of cable bends, joints, and surrounding obstacles on the complexity of the signal propagation path, compensate and correct the signal propagation speed, optimize the signal propagation model, and define a positioning objective function. Use an optimization algorithm to solve the minimum value of this objective function to obtain the coordinate of the insulation hidden danger position.

7. The method according to claim 6, characterized in that, In the step S13, it includes: When establishing a mathematical model for signal propagation, assume that there are a total of M sensors, and the amplitude of the signal collected by the i-th sensor is A i , and the phase is The distance from the signal source to the i-th sensor is d i , and the signal propagation speed is v; Calculate the signal propagation delay time t i , and the formula is: According to the signal propagation theory, the relationship between the signal amplitude and the propagation distance is expressed as: where A0 is the signal source amplitude, α is the coefficient related to the cable attenuation characteristic, and the relationship between the phase and the distance is: where ω is the signal angular frequency, is the initial phase of the signal source; Construct the positioning objective function J as: Use the gradient descent method to solve the minimum value of the objective function. Let the initial estimate of the coordinates of the insulation hidden danger be x 0 , set the learning rate η, the iteration number threshold T, and the convergence accuracy ∈, and take the partial derivative of the objective function J with respect to the coordinate variable x k to obtain the gradient vector as follows: In each iteration \(n = 0, 1, 2, \cdots\), update the estimated value of the insulation hidden danger position coordinates according to the update formula of the gradient descent method. The formula is: After each iteration, compare the objective function value J(x n ) of the current iteration with the objective function value J(x n-1 ) of the previous iteration. If |J(x n ) - J(x n-1 )| < ∈, it is considered that the algorithm converges. When the algorithm converges, obtain the current estimated coordinates of the insulation hidden danger location as the coordinates of the insulation hidden danger location to be solved and output them.

8. The method according to claim 7, characterized in that The step S13 further includes: Compensate and correct the signal propagation speed by evaluating the influence of the number of cable bends, joints, and surrounding obstacles on the complexity of the signal propagation path: Let the number of cable bends be B, the number of joints be J, the number of surrounding obstacles be O, and the signal propagation path complexity coefficient be C path The calculation formula for path is: C = α1B + α2J + α3O, Among them, α1, α2, and α3 are the corresponding influence weights. According to the value of C path , more targeted corrections are made to signal propagation. When C path is large, a compensation term is added when calculating the signal propagation speed v: v = v0(1 - βC path ), where v0 is the theoretical speed without the influence of complex paths, and β is the compensation coefficient.

9. The method according to claim 8, wherein In the step S14, it includes: Adjust the warning threshold according to the dynamic adjustment formula of the warning threshold. Let the cable operation time be t run , the initial warning threshold is T init , the aging coefficient is β, and the dynamic warning threshold T dynamic The calculation formula is: T dynamic = T init ·(1 + β·t run ).

10. The method according to claim 9, characterized in that, In the step S14, it includes: Dynamically adjust the warning threshold in combination with the real-time weather conditions. Let the real-time weather condition parameter vector be W = [w1, w2, …, w q , and the weather influence coefficient vector be k w = [k w1 , k w2 , …, k wq . The warning threshold is adjusted to:

Images