Early warning method and system for multiple hidden dangers of line

Abstract

The invention discloses a line multi-position hidden danger early warning method and system, and the method comprises the steps: obtaining first discharge data, and carrying out the first preprocessing of the first discharge data, and obtaining second discharge data; pre-training a first identification model; and establishing a first decomposition early warning strategy about a plurality of hidden danger types, and performing hidden danger early warning according to the output of the first identification model. The hidden danger detection accuracy is improved, and the potential hidden danger type can be identified more accurately through the pre-trained identification model. Automation of hidden danger early warning is achieved, the system can automatically conduct early warning according to output of the identification model, and the requirement for manual intervention is reduced. And through real-time monitoring and analysis of the discharge data, quick response can be realized, and an early warning signal can be sent out. The overall operation and maintenance efficiency is improved, and the deployment of the early warning system is helpful for taking measures in advance, so that the influence caused by faults is avoided or reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of early warning of hidden dangers at multiple locations on a line, and in particular to an early warning method and system for hidden dangers at multiple locations on a line. Background Art

[0002] Transmission and distribution lines are prone to insulator degradation, contamination, floating foreign objects, etc., which can lead to weakened insulation strength between the high-voltage side conductors and the ground wires and towers at the ground potential, causing hidden discharges. If not cleared in time, it will develop into a fault and eventually cause the line to trip and stop operating, resulting in significant economic losses and safety risks.

[0003] In order to accurately and timely identify and eliminate discharges on the line, it is necessary to first identify the existence and type of hidden discharges on the line. Currently, most transmission and distribution lines of various voltage levels have installed distributed traveling wave fault locating devices, which collect high-frequency discharge traveling waves generated and propagated from the discharge point at both ends of the line, calculate their characteristics, and compare and determine whether they have abnormal characteristics, so as to identify the existence and type of discharge, and notify the line operation and maintenance personnel to carry out maintenance work.

[0004] At present, similar technologies mainly extract high-frequency discharge waveform fragments to solve the problem of hidden danger discharge identification and classification, calculate specified features such as amplitude, phase, power frequency accompanying periodicity, etc., use pattern classification algorithms such as decision trees, support vector machines, neural networks, etc., and finally output the existence or type label of hidden dangers. However, in reality, it is very likely that hidden dangers of transmission lines occur in multiple places at the same time. When the transmission line crosses mountains and there is insulator deterioration in one place, there are trees that are too high in another place, or even multiple places. The high-frequency discharge waveforms they generate propagate in the line, affecting the diagnosis and location of hidden dangers. Therefore, the existing technology extracts a single or several waveform fragments and calculates features. This method does not take into account the development and accumulation of abnormal hidden dangers in the line. It is impossible to comprehensively judge the hidden danger status of the line from the entire process and predict future development trends. Therefore, it is easy to produce false alarms and missed alarms. Summary of the invention

[0005] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a method and system for early warning of hidden dangers at multiple locations on a line, which can solve the problems mentioned in the background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a method for early warning of multiple hidden dangers in a line, comprising:

[0010] Acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data;

[0011] Pre-train the first recognition model;

[0012] The input of the first identification model is the second discharge data, and the output is a number of hidden danger types;

[0013] A first decomposition warning strategy for several types of hidden dangers is established, and hidden danger warnings are performed according to the output of the first identification model.

[0014] As a preferred solution of the method for early warning of multiple hidden dangers of a line according to the present invention, the first identification model includes:

[0015] The first identification model is an arbitrary model that takes the second discharge data as input and outputs a number of hidden danger types or can directly or indirectly obtain hidden danger types.

[0016] As a preferred solution of the method for early warning of multiple hidden dangers in a line according to the present invention, the first decomposition early warning strategy includes:

[0017] Preset several risk levels and configure judgment thresholds for all risk levels;

[0018] Selecting characteristic parameters of the output of the first identification model;

[0019] A second preprocessing is performed on the characteristic parameter, and a risk level corresponding to the second preprocessing result is determined.

[0020] As a preferred solution of the method for early warning of multiple hidden dangers in lines described in the present invention, the second preprocessing at least includes historical data matching of the characteristic parameters and a first trend analysis.

[0021] As a preferred solution of the method for early warning of multiple hidden dangers in lines of the present invention, the second preprocessing further includes:

[0022] Perform a first trend analysis on the characteristic parameters after matching the historical data;

[0023] The first trend analysis includes dividing into several time periods;

[0024] Obtaining the discharge amount and the number of discharges in the plurality of time periods;

[0025] The risk level corresponding to the second preprocessing result is determined according to the discharge amount and the number of discharges in combination with the determination threshold.

[0026] As a preferred solution of the method for early warning of multiple hidden dangers of a line described in the present invention, the hidden danger types include at least tree discharge, insulator contamination discharge, insulator deterioration discharge, bird nest discharge and hanging object discharge.

[0027] As a preferred solution of the method for early warning of multiple hidden dangers in lines of the present invention, the first preprocessing includes:

[0028] Dividing the first discharge data according to cycles to establish a first data set;

[0029] The first data set includes several groups of test samples;

[0030] Performing half-wave information analysis on the first data set;

[0031] The half-wave information analysis includes at least the sequence number, data point size, maximum value and cumulative value of data points of each half-wave file.

[0032] In a second aspect, the present invention provides a multi-location hidden danger early warning system for a line, comprising:

[0033] A data preprocessing module, used for acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data;

[0034] A model building module, used for pre-training a first recognition model;

[0035] The input of the first identification model is the second discharge data, and the output is a number of hidden danger types;

[0036] The early warning module is used to establish a first decomposition early warning strategy for several types of hidden dangers and to issue a hidden danger early warning according to the output of the first identification model.

[0037] In a third aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method when executing the computer program.

[0038] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the method described above when executed by a processor.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention proposes a method and system for early warning of hidden dangers at multiple locations on a line, obtains first discharge data, and performs a first preprocessing on the first discharge data to obtain second discharge data; pre-trains a first identification model; the input of the first identification model is the second discharge data, and the output is several types of hidden dangers; establishes a first decomposition early warning strategy for several types of hidden dangers, and performs hidden danger early warning according to the output of the first identification model. The accuracy of hidden danger detection is improved, and the potential hidden danger types can be identified more accurately through the pre-trained identification model. The automation of hidden danger early warning is realized, and the system can automatically issue early warnings based on the output of the identification model, reducing the need for manual intervention. The timeliness of the early warning is enhanced, and through real-time monitoring and analysis of discharge data, it can quickly respond and issue early warning signals. The overall operation and maintenance efficiency is improved, and the deployment of the early warning system helps to take measures in advance to avoid or reduce the impact of failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0041] Figure 1 A method flow chart of a method and system for early warning of multiple hidden dangers in a line provided by an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of the 1st to 100th half-waves of a method and system for early warning of multiple hidden dangers in a line provided by an embodiment of the present invention;

[0043] Figure 3 A schematic diagram of characteristic parameter samples and characteristic extraction graph examples of a method and system for early warning of multiple hidden dangers in a line provided by an embodiment of the present invention;

[0044] Figure 4 A flowchart of a neural network classification implementation of a method and system for early warning of multiple hidden dangers in a line provided by an embodiment of the present invention;

[0045] Figure 5 A pollution type training mean square error (mse) trend change diagram of a line multiple hidden danger warning method and system provided by an embodiment of the present invention;

[0046] Figure 6 A general flow chart of a method and system for early warning of hidden dangers at multiple locations on a line provided by an embodiment of the present invention;

[0047] Figure 7A method for early warning of multiple hidden dangers in a line and a discharge quantity statistical trend distribution diagram of the system provided by an embodiment of the present invention;

[0048] Figure 8 A discharge pulse frequency distribution diagram of a method and system for early warning of multiple hidden dangers in a line provided by an embodiment of the present invention;

[0049] Fig. 9 An internal structural diagram of a computer device of a method and system for early warning of multiple hidden dangers in lines provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0051] Example 1

[0052] Reference Figure 1-Figure 9 , which is the first embodiment of the present invention, and provides a method and system for early warning of multiple hidden dangers in a line, including:

[0053] There are some problems in the existing related technologies, such as inaccuracy in data collection, delay in early warning response, and complexity of system maintenance.

[0054] The present application provides a method that can effectively solve the above-mentioned problems. Next, multiple embodiments will be combined to explain in detail how to implement the method for early warning of multiple hidden dangers on the line;

[0055] Figure 1 A method flow chart of a method and system for early warning of multiple hidden dangers in a line is shown, including:

[0056] S101, acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data;

[0057] In an optional embodiment, the first discharge data may be discharge waveform data collected in real time by a line monitoring device. The first preprocessing step includes filtering the original discharge waveform data to remove noise interference, and judging whether there is an abnormal discharge event by a set threshold.

[0058] In an alternative embodiment, the second discharge data are characteristic parameters extracted through a specific algorithm after filtering processing, and these parameters can reflect the health status of the line. Through such preprocessing and feature extraction, the early warning system can more accurately identify potential hidden dangers of the line, thereby sending out early warning signals in advance to avoid possible faults.

[0059] In an alternative embodiment, the first discharge data may include various parameters in the line, such as voltage, current, temperature, etc., and these parameters can comprehensively reflect the operating state of the line. By monitoring the changes of these parameters, the early warning system can more comprehensively evaluate the health status of the line. In addition, the system can also combine environmental factors, such as external conditions like humidity and temperature, to further improve the accuracy and reliability of the early warning. After data acquisition, the system will perform real-time analysis on these parameters. Once an anomaly is detected, the system will automatically activate the early warning mechanism and notify the maintenance personnel in a timely manner for inspection and handling, thereby effectively preventing the occurrence of line faults.

[0060] In an alternative embodiment, the first preprocessing may include filtering the collected line parameters to remove noise and interference and ensure the accuracy of the data. In addition, the preprocessing step may also include normalizing the data to make different parameters comparable, facilitating subsequent analysis and processing. Through these preprocessing steps, the early warning system can obtain higher-quality data input, thereby improving the early warning performance of the entire system.

[0061] In the embodiment of the present application, the first preprocessing includes:

[0062] Dividing the first discharge data by cycle to establish a first data set;

[0063] The first data set includes several groups of test samples;

[0064] Performing half-wave information analysis on the first data set;

[0065] The half-wave information analysis includes at least the serial number, data point size, maximum value, and cumulative value of the data points of each half-wave file.

[0066] Exemplarily, first, 500 * 2 cycles of discharge data are selected to obtain a discrete data set of 10 groups of test samples, and perform half-wave information analysis on it. Each group of test samples has 50 * 2 = 100 half-wave discharge data information, including the serial number, data point size, maximum value, and cumulative value of the data points of each half-wave file.

[0067] 1) Half-wave information analysis

[0068] According to the data sampling frequency of 5MHz, the half-wave time is 10ms, and the half-wave contains 10*5*103=5000 points. Based on the basic information of each half-wave file and the number of data points contained in each half-wave file, the number of half-waves composed of different half-wave file data, i.e., the half-wave division points, can be calculated.

[0069] Based on the half-wave split point, the half-wave information in all half-wave files is parsed to obtain the following information.

[0070]

[0071] 2) Half-wave characteristic statistics

[0072] Starting from 1, for every 100 consecutive half waves, a comparison curve of power frequency voltage and low-pass wave current is drawn, as shown below Figure 2 It is used to observe the half-wave files corresponding to the start and end of discharge, as well as the magnitude of abnormal discharge amplitude, etc.

[0073] In the process of drawing the original image, the half-wave power frequency voltage and the line low current are input into the half-wave feature extraction algorithm to realize the summary of the half-wave feature extraction results.

[0074]

[0075] Figure 3 The first figure in the feature extraction spectrum shows the distribution of pulse polarity and total number of pulses with a maximum discharge pulse amplitude of more than 50%. The left vertical axis represents the number of pulses (pieces), and it ranges from 0 to 400 with an interval of 100, and the horizontal axis is the number of half cycles (pieces), and it ranges from 0 to 1000 with an interval of 100; the second figure shows the distribution of maximum pulse values, the left vertical axis represents the current (A), and it ranges from 0 to 6 with an interval of 2, and the horizontal axis is the number of half cycles (pieces), and it ranges from 0 to 1000 with an interval of 100; the third figure shows the distribution of pulse energy with a maximum discharge pulse amplitude of more than 50%, the left vertical axis is the pulse energy (KWh), and it ranges from 0 to 2*10 -5 , the interval is 0.5*10 -5 The horizontal axis is the number of half cycles (pcs), and it ranges from 0 to 1000, with an interval of 100; the fourth figure shows the average rising edge time of the maximum discharge pulse amplitude above 50%, and the vertical axis on the left side shows the rising edge time (s), from 0 to 1*10 -5 , the interval is 0.5*10 -5 , the horizontal axis is the number of half cycles (pieces), and it ranges from 0 to 1000 with an interval of 100; the fifth figure shows the pulse phase distribution of the maximum discharge pulse amplitude of more than 50%, the values ​​of the vertical axis on the left are -100, 0, 100, and the horizontal axis is the number of half cycles (pieces), and it ranges from 0 to 1000 with an interval of 100;

[0076] It should be noted that obtaining the first discharge data and performing the first preprocessing on the first discharge data to obtain the second discharge data can effectively remove noise interference and improve the accuracy and reliability of the data. Through preprocessing, it can ensure that the discharge data for subsequent analysis is more stable, thereby providing a more accurate judgment basis for hidden danger warning.

[0077] S102, pre-training a first recognition model;

[0078] In the embodiment of the present application, the input of the first identification model is the second discharge data, and the output is a number of hidden danger types;

[0079] In the embodiment of the present application, the first identification model includes:

[0080] The first identification model takes the second discharge data as input and outputs a number of hidden danger types or any model that can directly or indirectly obtain hidden danger types.

[0081] In an optional embodiment, the first identification model can be constructed using a convolutional neural network (CNN), which can automatically extract features from the discharge waveform. Through training, CNN can identify waveform features corresponding to different types of hidden dangers, thereby accurately classifying line hidden dangers. In addition, in order to further improve the generalization ability of the model, data enhancement techniques can be used, such as rotating, scaling, translating, and other operations on waveform data to increase the diversity of training samples. In practical applications, other machine learning algorithms, such as support vector machines (SVM) or random forests (RF), can also be combined to improve the accuracy and robustness of the early warning system. Through these methods, it can be ensured that the early warning system can still maintain efficient hidden danger identification capabilities when faced with complex and changeable discharge data.

[0082] In an optional embodiment, the first recognition model can also be constructed using a recurrent neural network (RNN), which is particularly suitable for processing sequence data, such as time series discharge waveform data. RNN can capture the dynamic characteristics of the discharge waveform over time, which is crucial for understanding the discharge process and identifying hidden danger patterns. When training RNN, variants such as long short-term memory networks (LSTM) or gated recurrent units (GRU) can be used. These variants can effectively solve the problem of gradient vanishing or exploding of traditional RNN on long sequence data. By combining the advantages of CNN and RNN, a more powerful hybrid model can be constructed, which can not only extract the static characteristics of the discharge waveform, but also understand its dynamic changes over time, thereby achieving more accurate hidden danger prediction in the early warning system.

[0083] In an optional embodiment, the first identification model can also be constructed using a BP neural network, which can learn the complex nonlinear relationship between input data and output results through a multi-layer perceptron structure. The BP neural network has unique advantages in dealing with nonlinear problems, and is particularly suitable for situations where the mapping relationship between features and hidden dangers is more complex. By adjusting the number of layers of the network and the number of neurons in each layer, the performance of the model can be optimized to adapt to discharge data of different complexities. In addition, training using the back propagation algorithm can effectively reduce the model prediction error and improve the accuracy of the early warning system. In practical applications, other machine learning algorithms, such as support vector machines (SVM) or random forests (RF), can also be combined to further enhance the generalization ability and prediction accuracy of the model.

[0084] In the embodiment of the present application, the hidden danger types include at least tree discharge, insulator contamination discharge, insulator deterioration discharge, bird nest discharge and hanging object discharge.

[0085] In an optional embodiment, the hidden danger type may also include other possible hidden danger types, such as animal contact discharge, equipment aging discharge, lightning strike discharge, etc. These hidden danger types can cover a variety of abnormal situations that may be encountered during line operation, so that the early warning system can more comprehensively identify and respond to various potential risks.

[0086] In the embodiment of the present application, after the hidden danger data feature extraction is completed, all feature parameters are input into the BP neural network algorithm model to realize the hidden danger type identification. Since the BP neural network model has completed sufficient sample learning and accumulation through a large amount of experimental discharge data, after the newly input discharge data is identified by the existing training samples in the BP neural network model, a hidden danger type digital code will be obtained. In order to obtain accurate identification results, the Euclidean distance between the digital code and the set code is calculated for each experiment, and the hidden danger type corresponding to the digital code with the smallest Euclidean distance is taken as the identification result. The specific implementation process is as follows:

[0087] (1) First, set the execution parameters of the hidden danger type identification model of the BP neural network, including sample parameters, number of network layers, number of neurons, initialization weights, learning rate, target error, maximum iteration cycle, etc.

[0088] ① Set sample parameters: According to the selection requirements of the source analysis object, it is planned to collect 80 groups of existing training samples for each type of hidden danger discharge, and store the 20 groups of test samples used in the experiment into the waveform database to verify the recognition effect of the identification model.

[0089] ②Set the number of network layers: a total of 3 layers are set, including input layer, output layer, and hidden layer. The input layer represents the characteristic parameters of the hidden danger discharge data; the output layer represents the hidden danger discharge type. The five hidden danger discharge types are used as the expected output of the BP neural network. Tree discharge, insulator contamination discharge, insulator deterioration discharge, bird nest discharge, and floating object discharge are assigned values ​​according to the coding table; the hidden layer can realize weight approximation and achieve recognition effect.

[0090] ③ Determine the number of neurons: The number of neurons in the input layer is determined according to the number of characteristic parameters of the discharge data. The input layer of this experiment is planned to be set to 6 neurons, corresponding to the rising edge time, discharge pulse phase distribution, discharge pulse energy, discharge pulse frequency, frequency center of gravity, and main frequency range of the characteristic parameters of the discharge data. The number of neurons in the output layer is set to 1. The selection of the number of neurons in the hidden layer is relatively complex and should be adjusted according to the training process. Too few numbers will affect the recognition accuracy, and too many numbers will lead to excessive network calculations and stack overflow. After comparing the number of hidden layer neurons in the previous sample training, this experiment is planned to be set to 20.

[0091] ④ Initialize weights: At the beginning, assign a random value between -1 and 1 to the network weight.

[0092] ⑤ Determine the learning rate: During the BP neural network training process, the system continuously adjusts the weight threshold. Reasonable learning rate settings can make the adjustment of weights and thresholds fast and accurate. According to multiple design training processes, the learning rate value is set to 0.01.

[0093] ⑥ Target error: used to determine whether the iterative operation is completed. When its value is less than the result of the operation, the recognition process ends. Otherwise, the network will continue to adjust the weights and thresholds. According to the multiple design training process, its value is set to 0.01.

[0094] ⑦ Maximum iteration cycle: When the number of iterations of the network is greater than this value, but still not less than the target error, the training can be terminated. According to the multiple design training process, its value is 5000.

[0095] Table 1 BP neural network parameter setting value table

[0096]

[0097] (2) To ensure the speed of model training, all input feature parameters are normalized to speed up the calculation. After obtaining the normalized feature parameters, the obtained feature parameter training samples are input into the matlab neural network for calculation. The implementation process of the entire neural network classification is as follows: Figure 4 shown.

[0098] The mean square error of each training result is continuously reduced during the multiple iterations of the sample as a whole. Each iteration will have a back propagation, and as the error between the output value and the expected value continues to decrease, the weights of the neurons between each layer are also continuously adjusted and finally meet the requirement that the calculation error is less than the set target error value, and the final output digital code is obtained. Compared with the expected digital code, the hidden danger type corresponding to the digital code with the smallest Euclidean distance is taken as the identification result, thereby achieving the classification effect. At present, after the identification algorithm identifies the hidden danger data of the discharge towers on the 12# tower and the 23# tower, the digital codes are the closest to the Euclidean distance of the insulator contamination code, so it is determined that the discharge type under the two towers is the contaminated discharge type, as shown in Table 2 and Figure 5 shown.

[0099] Table 2 BP training recognition results

[0100]

[0101] It should be noted that the pre-trained first identification model can significantly improve the data processing speed and accuracy. Through pre-training, the model can learn the preliminary features of the discharge tower, thereby reducing the amount of calculation and improving the recognition efficiency in the subsequent feature extraction and classification tasks. In addition, the introduction of the pre-trained model enables the algorithm to adapt and adjust faster when facing new discharge types, thereby shortening the training time of the model and improving the generalization ability of the model.

[0102] S103, establishing a first decomposition warning strategy for several hidden danger types, and performing hidden danger warning according to the output of the first identification model.

[0103] In an optional embodiment, the first decomposition warning strategy can be designed in different ways to achieve rapid response to different types of hidden dangers. For example, different thresholds can be set to distinguish different levels of hidden dangers, and when the detected half-wave feature exceeds a specific threshold, the system will automatically trigger a corresponding warning signal.

[0104] In another optional embodiment, the probability of hidden dangers can be predicted based on historical data and statistical analysis, and the sensitivity of the early warning strategy can be adjusted accordingly to achieve the purpose of neither over-warning nor missing important hidden dangers. In this way, the early warning system can more accurately identify and respond to potential line hidden dangers, thereby improving the safety and reliability of the entire power system.

[0105] In an optional embodiment, the first decomposition warning strategy can also optimize the setting of the warning threshold by integrating a machine learning algorithm. Through the training data set, the algorithm can learn the relationship between different hidden danger types and half-wave characteristics, and automatically adjust the warning threshold to adapt to the actual operating conditions of the line. This method not only improves the accuracy of the warning, but also reduces the need for manual intervention, making the warning system more intelligent and automated.

[0106] In the embodiment of the present application, the first decomposition warning strategy includes:

[0107] Preset several risk levels and configure judgment thresholds for all risk levels;

[0108] Selecting characteristic parameters of the output of the first identification model;

[0109] A second preprocessing is performed on the characteristic parameter, and the risk level corresponding to the second preprocessing result is determined.

[0110] In the embodiment of the present application, the second preprocessing includes at least historical data matching of the characteristic parameters and a first trend analysis.

[0111] In an optional embodiment, the first trend analysis is used to analyze the changing trend of the characteristic parameters over time to predict the development trend of hidden dangers. Through the comparative analysis of historical data, the long-term trend and periodic changes of the characteristic parameters can be identified, thereby providing a more scientific basis for the early warning strategy. For example, if a certain characteristic parameter is found to show an upward trend in multiple consecutive cycles, the early warning system can issue an alarm in advance to prompt maintenance personnel to conduct inspections and maintenance to prevent the hidden danger from further developing into an actual failure.

[0112] In the embodiment of the present application, the second preprocessing further includes:

[0113] Perform a first trend analysis on the characteristic parameters after matching the historical data;

[0114] The first trend analysis involves dividing into several time periods;

[0115] Obtaining the discharge amount and the number of discharges in several time periods;

[0116] The risk level corresponding to the second preprocessing result is determined according to the discharge amount and the number of discharges in combination with the determination threshold.

[0117] In summary, the present invention proposes a method for early warning of multiple hidden dangers in a circuit, which obtains first discharge data and performs first preprocessing on the first discharge data to obtain second discharge data; pre-trains a first identification model; the input of the first identification model is the second discharge data, and the output is several hidden danger types; a first decomposition early warning strategy for several hidden danger types is established, and early warning of hidden dangers is carried out according to the output of the first identification model. The accuracy of hidden danger detection is improved, and potential hidden danger types can be more accurately identified through the pre-trained identification model. The automation of early warning of hidden dangers is realized, and the system can automatically give early warning according to the output of the identification model, reducing the need for manual intervention. The timeliness of early warning is enhanced, and through real-time monitoring and analysis of discharge data, early warning signals can be quickly responded to and sent out. The overall operation and maintenance efficiency is improved, and the deployment of the early warning system helps to take measures in advance to avoid or reduce the impact of faults.

[0118] Embodiment 2

[0119] In a preferred embodiment, the second preprocessing is used to perform a first trend analysis on the characteristic parameters after matching the historical data;

[0120] The first trend analysis includes dividing several time periods;

[0121] Obtain the discharge amount and the number of discharges for several time periods;

[0122] According to the discharge amount and the number of discharges, combine the judgment threshold to judge the risk level corresponding to the result of the second preprocessing.

[0123] In an alternative embodiment, for the time-domain characteristic change of pollution discharge, hierarchical early warning can be further realized. From the in-depth study of the early warning of various types of insulator hidden dangers in transmission lines, it can be obtained that the characteristic parameters for realizing the hierarchical early warning of polluted insulators are mainly realized from three characteristic quantities: the number of discharge pulses, pulse energy, and pulse distribution phase. Therefore, for these three characteristic quantities, historical trend change curves are made, and the process is as follows Figure 6 As shown, perform historical data trend analysis on the risk tower. If the tower does not reach the I-level risk standard, continue to monitor and count. If it reaches, hierarchical early warning (level one, level two, level three) will be carried out according to the risk classification standard at this voltage level.

[0124] Furthermore, the abscissa is defined as the time period, the ordinate is the discharge amount, and the statistical trend distribution diagram of the discharge amount is referred to as Figure 7 :

[0125] Furthermore, plot the number of half-cycle pulse discharges in each time period. The abscissa is the same as before, and the ordinate is the number of discharges, as Figure 8 shown:

[0126] It should be noted that since the insulator type of this line is glass insulator, combined with the graded warning judgment standards for glass insulators, it is determined that the insulator here is at the third level of contamination discharge risk.

[0127] Example 3

[0128] This embodiment also provides a line multiple hidden danger early warning system, including:

[0129] A data preprocessing module, used for acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data;

[0130] A model building module, used for pre-training a first recognition model;

[0131] The input of the first identification model is the second discharge data, and the output is a number of hidden danger types;

[0132] The early warning module is used to establish a first decomposition early warning strategy for several types of hidden dangers and to issue a hidden danger early warning according to the output of the first identification model.

[0133] The above-mentioned unit modules may be embedded in or independent of the processor in the computer device in the form of hardware, or may be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0134] This embodiment also provides a computer device, which may be a terminal, and its internal structure diagram may be as shown in FIG. Fig. 9 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for early warning of multiple hidden dangers in a line is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0135] This embodiment further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0136] Acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data;

[0137] Pre-train the first recognition model;

[0138] The input of the first identification model is the second discharge data, and the output is a number of hidden danger types;

[0139] A first decomposition warning strategy for several types of hidden dangers is established, and hidden danger warnings are performed according to the output of the first identification model.

[0140] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0141] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt 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.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0142] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0143] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0144] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0145] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0146] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for early warning of multiple hidden dangers in a line, characterized in that: include: Acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data; Pre-train the first recognition model; The input of the first identification model is the second discharge data, and the output is a number of hidden danger types; A first decomposition warning strategy for several types of hidden dangers is established, and hidden danger warnings are performed according to the output of the first identification model.

2. The method for early warning of multiple hidden dangers of a line according to claim 1, characterized in that: The first identification model includes: The first identification model is an arbitrary model that takes the second discharge data as input and outputs a number of hidden danger types or can directly or indirectly obtain hidden danger types.

3. The method for early warning of multiple hidden dangers of a line according to claim 2, characterized in that: The first decomposition warning strategy includes: Preset several risk levels and configure judgment thresholds for all risk levels; Selecting characteristic parameters of the output of the first identification model; A second preprocessing is performed on the characteristic parameter, and a risk level corresponding to the second preprocessing result is determined.

4. The method for early warning of multiple hidden dangers of a line according to claim 3, characterized in that: The second preprocessing at least includes historical data matching of the characteristic parameters and a first trend analysis.

5. The method for early warning of multiple hidden dangers of a line according to claim 4, characterized in that: The second preprocessing further comprises: Perform a first trend analysis on the characteristic parameters after matching the historical data; The first trend analysis includes dividing into several time periods; Obtaining the discharge amount and the number of discharges in the plurality of time periods; The risk level corresponding to the second preprocessing result is determined according to the discharge amount and the number of discharges in combination with the determination threshold.

6. The method for early warning of multiple hidden dangers of a line according to claim 5, characterized in that: The types of hidden dangers include at least tree discharge, insulator contamination discharge, insulator deterioration discharge, bird nest discharge and floating object discharge.

7. The method for early warning of multiple hidden dangers of a line according to claim 6, characterized in that: The first preprocessing comprises: Dividing the first discharge data according to cycles to establish a first data set; The first data set includes several groups of test samples; Performing half-wave information analysis on the first data set; The half-wave information analysis includes at least the sequence number, data point size, maximum value and cumulative value of data points of each half-wave file.

8. A multi-location hidden danger early warning system for a line, characterized in that: include: A data preprocessing module, used for acquiring first discharge data, and performing first preprocessing on the first discharge data to obtain second discharge data; A model building module, used for pre-training a first recognition model; The input of the first identification model is the second discharge data, and the output is a number of hidden danger types; The early warning module is used to establish a first decomposition early warning strategy for several types of hidden dangers and to issue a hidden danger early warning according to the output of the first identification model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images