A Fault Location Method for Overhead Hybrid Distribution Lines

Abstract

The present application discloses a fault location method for an overhead hybrid distribution line, which relates to the technical field of line fault location. The method includes: collecting monitoring data of the overhead hybrid distribution line, and estimating an initial fault area; dividing the initial fault area into grids to obtain a plurality of grid areas, and taking a plurality of grid area points in the grid areas; collecting area data of the grid area points, evaluating the abnormal probability of the grid area points, and selecting area measurement points; obtaining environmental data of the area measurement points, and evaluating the environmental interference degree of different variables; estimating the fault type through the monitoring data, obtaining the sensitivity of different variables to the fault type, collecting the measurement difficulty of the variables, and selecting the variables to be measured in combination with the environmental interference degree; measuring the variable data to obtain the fault location result. The present application improves the convenience of fault location for overhead hybrid distribution lines.

Description

Technical Field

[0001] This application relates to the technical field of line fault location, and in particular to a fault location method for overhead hybrid distribution lines. Background Art

[0002] An overhead hybrid distribution line refers to a composite power transmission system composed of overhead lines and cable segments. Its core advantage lies in combining the characteristics of the two types of lines to achieve a balance among economy, reliability, and scenario adaptability. At present, overhead hybrid distribution lines are widely used. However, due to the combination of the two types of lines, their coverage is extensive. When a fault occurs in the power transmission system, it is necessary to locate the fault in a timely manner to repair the fault and reduce the impact caused by the fault. Currently, the fault location of distribution lines is achieved by measuring a certain electrical variable or magnetic variable, and the measurement points are set randomly. The fault location result is obtained through measurements at multiple measurement points. On the one hand, randomly selecting measurement points will increase the measurement work. On the other hand, measuring a single variable will increase the difficulty and trouble of the measurement work at some measurement points, resulting in inconvenience in the fault location of distribution lines. Summary of the Invention

[0003] The purpose of the present invention is to provide a fault location method for overhead hybrid distribution lines to solve the problems raised in the above background art.

[0004] A fault location method for overhead hybrid distribution lines provided by this application adopts the following technical solutions:

[0005] Collect the monitoring data of the monitoring system of the overhead hybrid distribution line, and estimate the initial fault area based on the monitoring data;

[0006] Divide the initial fault area into grids to obtain multiple grid areas, and take multiple points in the grid areas as grid area points;

[0007] Collect the area data of the grid area points, evaluate the abnormal probability of the grid area points based on the area data, and select the area measurement points according to the abnormal probability;

[0008] Obtain the environmental data of the area measurement points, and evaluate the environmental interference degree of different variables based on the environmental data;

[0009] Estimate the fault type through the monitoring data, obtain the sensitivity of different variables to the fault type, collect the measurement difficulty of the variables, and select the variables to be measured in combination with the environmental interference degree;

[0010] Measure the variable data, perform normalization processing on the variable data, extract the variable change degree, and perform location based on the variable change degree to obtain the fault location result.

[0011] Preferably, the steps of collecting the regional data of the grid area points, evaluating the abnormal probability of the grid area points based on the regional data, and selecting the regional measurement points based on the abnormal probability are specifically as follows:

[0012] Collect the regional data of the grid area points, where the regional data includes environmental data and biological data;

[0013] Statistically analyze the change differences of different environmental factors in the environmental data, and superimpose the change differences of all environmental factors to obtain the environmental change value;

[0014] Obtain the line position of the overhead hybrid distribution line at the grid area points, and statistically analyze the presence rates of different organisms at the line positions;

[0015] Select the organisms with the presence rates reaching the preset presence rate threshold as the reflecting organisms, and extract the reflecting biological data from the biological data;

[0016] Evaluate the change situation of the reflecting organisms based on the reflecting biological data to obtain the biological change value, combine the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line, and select the grid area point with the maximum abnormal probability as the regional measurement point.

[0017] Preferably, the steps of evaluating the change situation of the reflecting organisms based on the reflecting biological data to obtain the biological change value, and combining the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line are specifically as follows:

[0018] Extract the change value of the habitat density of the reflecting organisms based on the reflecting biological data, and determine whether the change value of the habitat density reaches the preset density change threshold. If it reaches the preset density change threshold, use the change value of the habitat density as the biological abnormal value;

[0019] If it does not reach the preset density change threshold, obtain the daily behavior characteristics of the reflecting organisms, and extract the real-time behavior characteristics of the reflecting organisms based on the reflecting biological data;

[0020] Compare the feature similarity between the daily behavior characteristics and the real-time behavior characteristics, and use the reciprocal of the feature similarity as the biological abnormal value;

[0021] After normalizing all the biological abnormal values, calculate the average value of the biological abnormal values of all the reflecting organisms as the biological change value;

[0022] Use the linear regression algorithm to fit the linear relationship between the environmental change value and the biological change value and the abnormal probability of the overhead hybrid distribution line, generate the change linear equation, and calculate the abnormal probability of the overhead hybrid distribution line according to the change linear equation.

[0023] Preferably, the steps of obtaining the environmental data of the regional measurement points and evaluating the environmental interference degrees of different variables based on the environmental data are specifically as follows:

[0024] Obtain the environmental data of the regional measurement points in real time through a distributed sensor network, where the environmental data includes temperature W, humidity S, wind speed F, electromagnetic noise energy Z, and geomagnetic disturbance index C;

[0025] Perform normalization processing and denoising processing on the collected environmental data to generate a standardized data set;

[0026] Establish an environmental interference degree equation based on the standardized data set:

[0027] where DB is the electrical variable interference degree, and CB is the magnetic variable interference degree, is the temperature-humidity coupling function, is the wind speed effect function, is the geomagnetic disturbance coefficient, is the wind-induced magnetic circuit offset, is the function of the influence of temperature on magnetic permeability, , , , , , are proportionality coefficients, p is the label of different electrical variables, and q is the label of different magnetic variables;

[0028] Calculate the environmental interference degrees of different variables according to the environmental interference degree equation.

[0029] Preferably, the steps of predicting the fault type through the monitoring data, obtaining the sensitivity of different variables to the fault type, collecting the measurement difficulty of the variables, and selecting the measured variables in combination with the environmental interference degree are specifically as follows:

[0030] Obtain the historical fault data of the fault type, and extract the average change values of different variables during the fault from the historical fault data;

[0031] Perform normalization processing on the average change values of different variables to obtain the basic sensitivities of different variables;

[0032] Obtain the line form of the overhead hybrid distribution line at the regional measurement point and record it as the measurement line segment, and the line form is divided into overhead line segments and cable line segments;

[0033] Obtain the resistivity and dynamic impedance fluctuation values of the overhead line segment, and combine the basic sensitivity of the electrical variable to obtain the variable sensitivity of the electrical variable of the overhead line segment, and use the basic sensitivity of the magnetic variable as the variable sensitivity of the magnetic variable of the overhead line segment;

[0034] Obtain the capacitance effect value of the cable line segment, and obtain the variable sensitivity of the electrical variable of the cable line segment according to the capacitance effect value and the basic sensitivity of the electrical variable;

[0035] Obtain the shielding performance value of the cable segment, and obtain the variable sensitivity of the magnetic variable of the cable segment according to the shielding performance value and the basic sensitivity value of the magnetic variable;

[0036] Obtain the variable sensitivity of different variables corresponding to the measurement segment and record it as the segment sensitivity, collect the measurement selection degree of the variable, and calculate the measurement degree of different variables in combination with the environmental interference degree, and select the variable with the largest measurement degree for measurement.

[0037] Preferably, the step of collecting the measurement selection degree of the variable is specifically:

[0038] Obtain the measurement method of the variable, judge whether the measurement method of the variable is single, if it is single, then judge whether the measured variable must be in physical contact with the measurement segment to obtain the contact judgment result;

[0039] Obtain the operation data for the measurement segment to satisfy physical contact, and obtain the variable measurement difficulty according to the operation data and the contact judgment result;

[0040] If the measurement method is not single, then use the minimum variable measurement difficulty in the measurement method as the variable measurement difficulty of the variable;

[0041] Evaluate the capture difficulty of the variable according to the historical failure data, obtain the known data, and evaluate the effectiveness of the variable according to the known data;

[0042] Combine the variable measurement difficulty, capture difficulty and effectiveness to obtain the measurement selection degree of the variable.

[0043] Preferably, the step of obtaining the operation data for the measurement segment to satisfy physical contact and obtaining the variable measurement difficulty according to the operation data and the contact judgment result is specifically:

[0044] If it must be in physical contact with the measurement segment, then count the operation duration, operation danger degree and operation influence degree according to the operation data;

[0045] Evaluate the basic operation difficulty of the variable according to the operation data;

[0046] Obtain the measurement steps of the variable, evaluate the basic measurement difficulty of the variable according to the measurement steps, and combine the basic operation difficulty to obtain the variable measurement difficulty;

[0047] If it is not necessary to be in physical contact with the measurement segment, then record the basic measurement difficulty as the variable measurement difficulty.

[0048] Preferably, the step of evaluating the capture difficulty of the variable according to the historical failure data is specifically:

[0049] Extract the average duration of different variable changes during the failure from the historical failure data;

[0050] Estimate the fault degree based on the monitoring data, and estimate the dynamic range of variable fluctuations in combination with the line segment sensitivity of the variable;

[0051] Judge whether the dynamic range is covered by a single measurement operation. If the dynamic range is covered by a single measurement operation, establish a correlation curve between the average duration and the capture difficulty, and find the corresponding capture difficulty according to the correlation curve;

[0052] If the dynamic range is not covered by a single measurement operation, obtain the fluctuation frequency of variable fluctuations, and obtain the capture difficulty according to the fluctuation frequency.

[0053] Preferably, the step of obtaining the fluctuation frequency of variable fluctuations and obtaining the capture difficulty according to the fluctuation frequency when the dynamic range is not covered by a single measurement operation is specifically as follows:

[0054] Obtain the measurement operation methods covering the dynamic range, and obtain the method coverage range of each measurement operation method;

[0055] Record the variable fluctuation from one method coverage range to another method coverage range as one fluctuation, and count the fluctuation frequency of the variable;

[0056] Count the average existence duration of variable fluctuations in different method coverage ranges, and obtain the capture difficulty in combination with the fluctuation frequency.

[0057] Preferably, the step of obtaining the known data and evaluating the validity of the variable according to the known data is specifically as follows:

[0058] Obtain the known data of the overhead hybrid distribution line, and count the calculation data by calculating the variable and the known data;

[0059] Screen the calculation data related to fault location as valid data, and find the variable sensitivity corresponding to the valid data;

[0060] Sum all the valid data and the corresponding variable sensitivities to obtain the validity of the variable.

[0061] In summary, the present application includes at least one of the following beneficial technical effects:

[0062] 1. An initial fault area is estimated based on the monitoring data of the overhead hybrid distribution line. The initial fault area is divided into multiple grid areas, and multiple grid area points are selected from them. According to the environmental change value and the biological change value obtained from the change in the habitat density of organisms and the change in the behavioral characteristics of organisms, the abnormal probability of the overhead hybrid distribution line at the grid area points is comprehensively evaluated, so as to select the grid area point with the highest abnormal probability to measure its electrical variable or magnetic variable. Selecting the point with the highest abnormal probability for measurement can, on the one hand, measure effective fault data to provide effective and obvious data for fault location, and on the other hand, can effectively reduce unnecessary measurements, reduce measurement blind spots, and improve the convenience of fault location of the overhead hybrid distribution line.

[0063] 2. The average change values of different variables during a fault are extracted from the historical fault data, and after being standardized, the basic sensitivity is obtained. Combining the influence of the line segment form of the overhead hybrid distribution line on the sensitivity, the corresponding variable sensitivity is obtained. Then, considering the measurement difficulty of the variable and the degree of interference of the environment on the variable calculated through the environmental interference degree equation, the measurement degree is obtained, and the variable with the largest measurement degree is selected for measurement. During the fault location process, selecting a more appropriate variable for measurement is beneficial to reducing the workload, locating the fault more quickly and accurately, and improving the accuracy of fault location of the overhead hybrid distribution line.

[0064] 3. The measurement difficulty of different measurement methods of the variable is evaluated according to the operation situation required for measurement and the line segment form of the overhead hybrid distribution line, and the minimum measurement difficulty is selected as the variable measurement difficulty. The capture difficulty of the variable is comprehensively evaluated according to the dynamic range of the variable fluctuation, the existence duration of the change, and whether the variable measurement operation is single. Then, combined with the effective data situation that can be obtained after the variable measurement, the measurement difficulty of the variable is comprehensively obtained, which is beneficial to selecting a simpler measurement variable, reducing the trouble of the staff, and improving the convenience of measuring variables in the fault location of the overhead hybrid distribution line. Description of the Drawings

[0065] Figure 1 It is a schematic diagram of the specific steps of an embodiment of a fault location method for an overhead hybrid distribution line of the present invention. Detailed Embodiment

[0066] The following combines the embodiments and Figure 1 to make a further detailed description of the present invention, but the embodiments of the present invention are not limited thereto.

[0067] The present invention discloses a fault location method for an overhead hybrid distribution line, which specifically includes the following steps:

[0068] Step S1, collect the monitoring data of the monitoring system of the overhead hybrid distribution line, and estimate the initial fault area according to the monitoring data.

[0069] In the prior art, overhead hybrid distribution lines are monitored, with several monitoring points. Based on the abnormal data uploaded by the monitoring points, the fault area can be initially located. For example, the system uses current and voltage sensors on the line to collect data in real time. When a sudden change in current or abnormal voltage fluctuation is detected at a certain location, a fault signal is triggered, and this location is used as the initial fault area.

[0070] Step S2: Divide the initial fault area into grids to obtain multiple grid areas, and select multiple points as grid area points in each grid area.

[0071] Randomly select multiple points on the overhead hybrid distribution lines in each grid area as grid area points, that is, each grid area has corresponding multiple grid area points.

[0072] Step S3: Collect the area data of the grid area points, evaluate the abnormal probability of the grid area points based on the area data, and select area measurement points according to the abnormal probability.

[0073] Evaluate the abnormal probability of multiple grid area points in each grid area, and then select the corresponding area measurement points in this grid area, that is, each grid area has corresponding area measurement points.

[0074] Step S4: Obtain the environmental data of the area measurement points, and evaluate the environmental interference degrees of different variables based on the environmental data.

[0075] Step S5: Estimate the fault type through the monitoring data, obtain the sensitivities of different variables to the fault type, collect the measurement difficulties of the variables, and select the variables to be measured in combination with the environmental interference degree.

[0076] Step S6: Measure the variable data, perform normalization processing on the variable data, then extract the variable change degree, and perform positioning according to the variable change degree to obtain the fault location result.

[0077] In actual operation, in an overhead hybrid distribution circuit, since it is a combination of two types of lines, the measurement of electrical variables or magnetic variables will be affected differently at different positions due to the different line forms. Selecting the points that are more likely to have faults for measurement is conducive to obtaining more effective data and reducing ineffective measurement work. Selecting the variables to be measured is conducive to reducing the trouble caused by measurement and more quickly and accurately locating the fault. After selecting the type of variable to be measured, measure the value of the variable. After normalizing the variable data at different measurement points, obtain the degree of change of the variable. Connect the points with the largest degree of change of the variable, and further perform step-by-step measurement on the line between the two points to obtain the fault location result. For example, measure the voltage at point A, measure the magnetic flux at point B, and measure the voltage at point C. After normalization, the degrees of change of the variables between points A, B, and C are 50%, 80%, and 30% respectively. Then the fault is located on the line between points A and B, and the AB line is measured point by point again, and the position with the largest degree of change of the variable is selected as the fault location.

[0078] The steps of collecting the regional data of the grid area points, evaluating the abnormal probability of the grid area points according to the regional data, and selecting the regional measurement points according to the abnormal probability are specifically as follows:

[0079] Step S31, collect the regional data of the grid area points, and the regional data includes environmental data and biological data.

[0080] The environmental data includes environmental factors such as temperature, humidity, and air content.

[0081] Step S32, count the change differences of different environmental factors in the environmental data, and superimpose the change differences of all environmental factors to obtain the environmental change value.

[0082] Set a time period, and record the difference between the environmental factors around the distribution line and the environmental factors in the distribution line area as the change difference. For example, at ten o'clock, the temperature is 23 degrees Celsius, and at ten thirty, the temperature is 24 degrees Celsius, and the change difference of the temperature is 1 degree Celsius. Before superimposing the change differences of different environmental factors, normalize the data. For example, the change difference range of the temperature is 1-10, the change difference range of the humidity is 1-20, the change difference of the temperature is 5 degrees Celsius, and the change difference of the humidity is 5RH%. After normalization, the numerical values of the change difference of the temperature and the change difference of the humidity are 10 and 5 respectively.

[0083] Step S33, obtain the line position of the overhead hybrid distribution line of the grid area points, and count the presence rates of different organisms at the line positions.

[0084] Due to different biological habits, different organisms stay or survive at different line locations. For example, there are mostly birds staying on overhead lines, while cable segments are located on the ground and the presence rate of birds will decrease significantly. And the probability of some organisms such as cats and dogs staying on the ground cable segments is greater. The presence rate refers to the probability of different organisms appearing at this line location.

[0085] Step S34, select the organisms with a presence rate reaching a preset presence rate threshold as the reflection organisms, and extract the reflection organism data from the biological data.

[0086] Some organisms only stay occasionally, so they cannot give effective data reflection on the line faults at this location. Therefore, selecting organisms with a higher presence rate can more effectively feedback whether the line has a fault.

[0087] Step S35, evaluate the change situation of the reflection organisms according to the reflection organism data to obtain a biological change value, combine it with the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line, and select the grid area point with the maximum abnormal probability as the area measurement point.

[0088] In practical applications, construct a data set containing the biological change value and the environmental change value, select the random forest model for training, use the trained model, and according to the input biological change value and environmental change value, output the prediction result of the abnormal probability of the overhead hybrid distribution line. The changes in the environment and organisms can reflect the sudden changes in the distribution line, thereby assisting in selecting measurement points and reducing unnecessary measurements. For example, when the line is short-circuited, the current surges and the wire resistance generates heat, so the temperature around the distribution line rises sharply. When there is a short-circuit or ground fault, the current mutates and the magnetic field intensity around it surges, and the abnormal magnetic flux may interfere with birds relying on geomagnetic navigation, forcing them to flee or change their flight paths.

[0089] The steps of evaluating the change situation of the reflection organisms according to the reflection organism data to obtain a biological change value and combining it with the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line are specifically as follows:

[0090] Step S351, extract the change value of the habitat density of the reflection organisms according to the reflection organism data, judge whether the change value of the habitat density reaches a preset density change threshold, and if it reaches the preset density change threshold, use the change value of the habitat density as the biological anomaly value.

[0091] Set a time period to observe the change situation of the habitat density of the organisms within the time period, that is, calculate the difference in habitat density within the time period as the change value of the habitat density. If the change value of the habitat density reaches the density change threshold, it means that this type of organism reduces the impact brought by the distribution line by fleeing, so it can be directly used as the biological anomaly value.

[0092] Step S352: If the preset density change threshold is not reached, obtain the daily behavior characteristics of the organisms, and extract the real-time behavior characteristics of the organisms based on the biological data.

[0093] If the preset density change threshold is not reached, the reflected phenomenon of the organisms being affected may be a change in behavior characteristics.

[0094] Step S353: Compare the feature similarity between the daily behavior characteristics and the real-time behavior characteristics, and take the reciprocal of the feature similarity as the biological outlier value.

[0095] For example, if the feature similarity is 90%, the biological outlier value is 1 / 90% = 1.11.

[0096] Obtain the feature similarity between the daily behavior characteristics and the real-time behavior characteristics through the cosine similarity model.

[0097] Step S354: After normalizing all the biological outlier values, calculate the average value of all the biological outlier values reflecting the organisms as the biological change value.

[0098] Normalization is a data standardization method that scales the data proportionally through mathematical transformation so that it falls into a small specific interval.

[0099] Step S355: Use the linear regression algorithm to fit the linear relationship between the environmental change value and the biological change value and the abnormal probability of the overhead hybrid distribution line, generate the change linear equation, and calculate the abnormal probability of the overhead hybrid distribution line according to the change linear equation.

[0100] In practical applications, the failure of the distribution line will cause changes in electrical variables or magnetic variables, and some organisms are sensitive to these variables. Therefore, the abnormal probability of the distribution line can be estimated through the abnormal behavior of the organisms. For example, a certain farmland study shows that after the line fails, the electric field intensity rises from the normal 0.5 kV / m to 3 kV / m, resulting in a 40% decrease in the bee returning-to-hive efficiency, the bee colony migrating to the area with a weaker electric field, a decrease in the local pollinator density, and thus a decrease in the habitat density, reflecting the abnormality of the distribution line. The short-circuit fault causes local distortion of the geomagnetic field (such as a 5° deviation of the magnetic declination), affecting the ability of birds (such as pigeons) to use magnetic field positioning. The number of times of calibrating the direction increases by 3 times, but the nest is not abandoned, and the breeding success rate decreases by 20% due to increased energy consumption, which also reflects the abnormality of the overhead hybrid distribution line.

[0101] The steps to obtain the environmental data of the regional measurement points and evaluate the environmental interference degrees of different variables according to the environmental data are specifically as follows:

[0102] Step S41: Obtain the environmental data of the regional measurement points in real time through a distributed sensor network. The environmental data includes temperature W, humidity S, wind speed F, electromagnetic noise energy Z, and geomagnetic disturbance index C.

[0103] Step S42: Perform normalization processing and denoising processing on the collected environmental data to generate a standardized data set.

[0104] Step S43: Establish an environmental interference degree equation based on the standardized data set:

[0105] where DB is the electrical variable interference degree, and CB is the magnetic variable interference degree. is the temperature-humidity coupling function. is the wind speed effect function. is the geomagnetic disturbance coefficient. is the wind vibration magnetic circuit offset. is the influence function of temperature on magnetic permeability. , , , , , are proportionality coefficients, p is the label of different electrical variables, and q is the label of different magnetic variables.

[0106] The temperature-humidity coupling function is calibrated by the non-linear relationship between the wire resistivity and temperature-humidity through laboratory accelerated aging experiments. For example, a polynomial fitting function is obtained according to the experimental data. The wind speed effect function is measured by the linear relationship between the wire dancing amplitude and the current volatility through a wind tunnel experiment. The geomagnetic disturbance coefficient is obtained by fitting an exponential function based on the geomagnetic station data and the measured values of the magnetic flux sensor, such as . The wind vibration magnetic circuit offset is calibrated by the linear slope of the wind speed and the magnetic circuit offset through a shaking table experiment. The influence function of temperature on magnetic permeability is determined by the iron magnetic material temperature rise experiment, such as .

[0107] Step S44: Calculate the environmental interference degrees of different variables according to the environmental interference degree equation.

[0108] In practical applications, the environment will interfere with electrical variables and magnetic variables. A high-temperature environment will increase the resistance value of wires or components, resulting in current measurement errors. For example, when the temperature of the meter acquisition device exceeds 50 °C, the internal resistance temperature drift may cause a data deviation of up to 5%. Since the environmental interferences on electrical variables and magnetic variables are different, two different interference degree models are generated and combined into an interference degree equation.

[0109] Steps for predicting the fault type through monitoring data, obtaining the sensitivity of different variables to the fault type, collecting the measurement difficulty of variables, and selecting the measured variables in combination with the environmental interference degree are as follows:

[0110] Step S51: Obtain the historical fault data of the fault type, and extract the average change values of different variables during the fault from the historical fault data.

[0111] The historical fault data refers to the relevant data recorded during the fault of the distribution line, including the real-time values of various electrical variables, magnetic variables, etc.

[0112] Step S52: Standardize the average change values of different variables to obtain the basic sensitivities of different variables.

[0113] The sensitivities of different variables to the fault type are different. For example, in the case of repeated reignition of the arc, the current and magnetic field oscillate at high frequencies. The change value of the electrical variable fluctuates from 1 - 2A under normal conditions to 5 - 20A, with a relatively small change. However, the voltage of the non-fault phase increases to 1.5 times the rated voltage, with a relatively large change.

[0114] Step S53: Obtain the line segment form of the overhead hybrid distribution line at the regional measurement point and record it as the measured line segment. The line segment form is divided into overhead line segments and cable line segments.

[0115] The line segment forms of different overhead hybrid distribution lines will cause changes in the sensitivity of variables to faults.

[0116] Step S54: Obtain the resistivity and dynamic impedance fluctuation value of the overhead line segment, and combine the basic sensitivity of the electrical variable to obtain the variable sensitivity of the electrical variable of the overhead line segment. Take the basic sensitivity of the magnetic variable as the variable sensitivity of the magnetic variable of the overhead line segment.

[0117] In the overhead line, the sensitivity of the electrical variable is inhibited by high resistivity and dynamic impedance fluctuations. However, due to the open structure, the magnetic variable performs excellently during the fault, and the sensitivity is less affected. Collect multiple sets of resistivity and dynamic impedance fluctuation values, and use non-linear regression technology to establish a non-linear equation with resistivity and dynamic impedance fluctuation values as independent variables. Calculate the cable sensitivity of the electrical variable of the overhead line segment through this equation. For example, overhead lines usually use aluminum stranded wire or steel-core aluminum stranded wire, and their resistivity is higher than that of copper-core cables. During a short-circuit fault, the inherent characteristics of the line impedance will inhibit the sudden change of current. For example, under the same fault conditions, the current increase of the copper-core cable may reach 10%, while for the overhead line, due to its higher resistivity, the actual increase may only be 5%. Calculate the average value of the basic sensitivity and the cable sensitivity as the variable sensitivity.

[0118] Step S55: Obtain the capacitance effect value of the cable line segment, and obtain the variable sensitivity of the electrical variable of the cable line segment according to the capacitance effect value and the basic sensitivity of the electrical variable.

[0119] The coaxial structure of the cable results in the capacitance to the ground per unit length being more than 10 times that of the overhead line. In the case of a high-resistance grounding fault, the capacitive current will mask the true change of the zero-sequence current, significantly reducing the sensitivity of electrical variables. Therefore, the capacitive effect affects the sensitivity of electrical variables. Establish the correlation curve between the capacitive effect value and the sensitivity, find the cable sensitivity according to the correlation curve, and then calculate the average value of the basic sensitivity and the cable sensitivity to obtain the variable sensitivity.

[0120] Step S56: Obtain the shielding performance value of the cable segment, and obtain the variable sensitivity of the magnetic variable of the cable segment according to the shielding performance value and the basic sensitivity value of the magnetic variable.

[0121] The sheath cancels the internal magnetic field through eddy currents, suppressing the propagation of magnetic signals of high-frequency arc faults. For example, the high-frequency magnetic field intensity of an internal arc fault may be attenuated by more than 50% outside the sheath, and the sensitivity of the magnetic variable decreases. The shielding performance value is obtained through parameter extraction of the cable segment, and is calculated by using the above method of obtaining the variable sensitivity from the capacitive effect value and the basic sensitivity.

[0122] Step S57: Obtain the variable sensitivity of the corresponding different variables according to the measured segment and record it as the segment sensitivity, collect the measurement selectivity of the variable, and calculate the measurement degree of different variables in combination with the environmental interference degree, and select the variable with the largest measurement degree for measurement.

[0123] In practical applications, a neural network can be designed and trained. Its input layer includes nodes of segment sensitivity, measurement selectivity, and environmental interference degree, and the output layer is the measurement degree. Use historical data to train the neural network and adjust the network weights to minimize the prediction error. Input the real-time segment sensitivity, measurement selectivity, and environmental interference degree nodes into the trained neural network to obtain the predicted output of the measurement degree.

[0124] The steps of collecting the measurement selectivity of the variable are specifically as follows:

[0125] Step S571: Obtain the measurement method of the variable, judge whether the measurement method of the variable is single, and if it is single, judge whether the measured variable must be in physical contact with the measured segment to obtain the contact judgment result.

[0126] There may be more than one measurement method for different variables. For example, to measure current, an ammeter can be directly connected in series in the circuit to be measured, or a clamp ammeter can measure the current by inducing the magnetic field around the wire to be measured.

[0127] Step S572: Obtain the operation data for the measured segment to meet physical contact, and obtain the variable measurement difficulty according to the operation data and the contact judgment result.

[0128] Step S573, if the measurement methods are not single, then use the measurement difficulty of the smallest variable in the measurement methods as the variable measurement difficulty of the variable.

[0129] Then obtain the measurement difficulty of each measurement method, and use the smallest measurement difficulty among them as the variable measurement difficulty. For example, there are three methods A, B, and C for current measurement, and the difficulty of method A is the smallest, so it is used as the measurement difficulty of current measurement.

[0130] Step S574, evaluate the capture difficulty of the variable according to the historical fault data, obtain the known data, and evaluate the validity of the variable according to the known data.

[0131] Step S575, combine the variable measurement difficulty, capture difficulty, and validity to obtain the measurement selection degree of the variable.

[0132] In practical applications, train the variable measurement difficulty, capture difficulty, and validity through a linear regression model to obtain the variable measurement difficulty. If there are multiple measurement methods for a variable, then use the smallest variable measurement difficulty as the variable measurement difficulty.

[0133] The steps to obtain the operation data where the measurement line segment satisfies physical contact and obtain the variable measurement difficulty according to the operation data and the contact judgment result are specifically as follows:

[0134] Step S5721, if physical contact with the measurement line segment is necessary, then count the operation duration, operation danger degree, and operation influence degree according to the operation data.

[0135] The operation duration is the average duration for measuring this variable. The operation danger degree can be obtained according to the accident incidence rate of measuring this variable, or can be evaluated by the user. The operation influence degree can be obtained through expert evaluation. The operation influence degree refers to the influence on the overhead hybrid distribution line, surrounding building facilities, etc. during the operation.

[0136] Step S5722, evaluate the basic operation difficulty of the variable according to the operation data.

[0137] Train the operation data through a multivariate non-linear regression model to obtain the basic operation difficulty. Since the operations before physical contact with the measurement line segment are the same for different variables, it is used as the basic operation difficulty.

[0138] Step S5723, obtain the measurement steps of the variable, evaluate the basic measurement difficulty of the variable according to the measurement steps, and combine the basic operation difficulty to obtain the variable measurement difficulty.

[0139] The basic measurement difficulty can be evaluated according to the number of measurement steps. The more the measurement steps, the greater the basic measurement difficulty. The basic measurement difficulty can also be obtained by evaluating the difficulty of the measurement steps by the user.

[0140] Step S5724: If physical contact with the measurement line segment is not necessary, record the basic measurement difficulty as the variable measurement difficulty.

[0141] In practical applications, when the measurement must be in physical contact with the line segment, overhead line segments require high-altitude operations, and underground cables require digging up the ground, both of which increase the measurement difficulty to a certain extent. The measurement of the variable itself also has a certain measurement difficulty. For example, when measuring the current by directly connecting an ammeter in series in the measured circuit, it is necessary to connect to the circuit, increasing the operation risk. If no contact is required, the previous operations are not needed, and the basic measurement difficulty can be directly used as the variable measurement difficulty. If physical contact is required, it is also necessary to combine the basic operation difficulty, and the variable measurement difficulty is obtained through linear regression model training.

[0142] The steps for evaluating the capture difficulty of the variable based on historical fault data are specifically as follows:

[0143] Step S5741: Extract the average duration of different variable changes during the fault from the historical fault data.

[0144] For example, when a certain fault occurs, the current surges to 100A, and the average duration of the current remaining at 100A.

[0145] Step S5742: Estimate the fault degree based on the monitoring data, and combine the line segment sensitivity of the variable to estimate the dynamic range of variable fluctuations.

[0146] Time-domain / frequency-domain features (such as mean value, variance, harmonic content) can be extracted from the existing monitoring data (such as current, voltage, temperature), and thresholds or trend thresholds are set to judge the fault degree, and the fault degree has its corresponding dynamic range.

[0147] Step S5743: Determine whether the dynamic range is covered by a single measurement operation. If the dynamic range is covered by a single measurement operation, establish a correlation curve between the average duration and the capture difficulty, and find the corresponding capture difficulty according to the correlation curve.

[0148] A single operation refers to not changing the measurement method, measurement equipment, or adjusting the measurement equipment, that is, no further adjustment is required during the measurement, which is a single measurement operation.

[0149] Step S5744: If the dynamic range is not covered by a single measurement operation, obtain the fluctuation frequency of the variable fluctuation, and obtain the capture difficulty according to the fluctuation frequency.

[0150] In practical applications, if the dynamic range can be covered by a single measurement operation, the capture difficulty comes from the average duration of the change. The longer the average duration, the easier it is to capture. For example, in some faults, the current changes instantaneously, making it difficult to capture in time.

[0151] If the dynamic range is not covered by a single measurement operation, the steps of obtaining the fluctuation frequency of the variable fluctuation and obtaining the capture difficulty according to the fluctuation frequency are as follows:

[0152] Step S57441: Obtain the measurement operation methods that cover the dynamic range, and obtain the coverage range of each measurement operation method.

[0153] Adjusting the operating device, replacing the operating device, etc. are all different measurement operation methods.

[0154] Step S57442: Record the variable fluctuation from the coverage range of one method to the coverage range of another method as one fluctuation, and count the fluctuation frequency of the variable.

[0155] For example, the coverage range of measurement operation method A is 1 - 10A, the coverage range of measurement operation method B is 5 - 20A, and the coverage range of measurement operation method C is 20 - 50A. If the current value fluctuates from 3A to 8A and remains within the coverage range of measurement operation method A, it is not considered a fluctuation. However, if the current value fluctuates from 7A to 18A, it is recorded as one fluctuation because it fluctuates from the coverage range of measurement operation method A to the coverage range of measurement operation method B. Calculate the relationship between the number of fluctuations and time to obtain the fluctuation frequency, such as 10 times / minute.

[0156] Step S57443: Statistically analyze the average existence duration of the variable fluctuation within the coverage ranges of different methods, and combine the fluctuation frequency to obtain the capture difficulty.

[0157] In practical applications, the weight ratios of the average existence duration and the fluctuation frequency are set respectively, and the capture difficulty is calculated according to the weight ratios. For example, the weight ratios of the average existence duration and the fluctuation frequency are 40% and 60% respectively, and the average existence duration and the fluctuation frequency are 40 minutes and 5 times / minute respectively. Then the capture difficulty is 40×40% + 5×60% = 19.

[0158] The steps of obtaining known data and evaluating the validity of variables according to the known data are as follows:

[0159] Step S5745: Obtain the known data of the overhead hybrid distribution line, and statistically analyze the variables and the known data to calculate the calculated data.

[0160] The calculated data refers to the data that has not been measured and can be calculated through common knowledge, existing formulas, etc.

[0161] Step S5746: Screen the calculated data related to fault location as valid data, and find the variable sensitivity corresponding to the valid data.

[0162] The calculated data related to fault location refers to the calculated data that is useful for fault location as valid data.

[0163] Step S5747: Sum all the valid data and the corresponding variable sensitivities to obtain the validity of the variable.

[0164] In practical applications, other variables can be calculated through an electrical variable or a magnetic variable, but some other variables are required for auxiliary calculation. For example, when measuring current, if the resistance is known, the voltage can be calculated through Ohm's law (V = IR). In fault location, the voltage can also play a role, so it is considered as valid data. However, the voltage cannot be calculated without knowing the resistance. Therefore, check whether there is a corresponding calculation formula based on the variable and the known data. If there is, use the result of the calculation formula as the calculated data. Screen out the data useful for fault location as the valid data. The variable sensitivity of the valid data is obtained through the above method of variable sensitivity. The greater the variable sensitivity, the greater the validity of the variable, indicating that the obtained data is more useful for fault location.

[0165] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited accordingly. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A fault location method for overhead hybrid distribution lines, characterized in that, It includes the following steps: Collect the monitoring data of the monitoring system for the overhead hybrid distribution line, and estimate the initial fault area based on the monitoring data; Divide the initial fault area into grids to obtain multiple grid areas, and take multiple points in the grid areas as grid area points; Collect the area data of the grid area points, evaluate the abnormal probability of the grid area points based on the area data, and select area measurement points according to the abnormal probability; Obtain the environmental data of the area measurement points, and evaluate the environmental interference degrees of different variables based on the environmental data; Estimate the fault type through the monitoring data, obtain the sensitivities of different variables to the fault type, collect the measurement difficulty of the variables, and select the variables to be measured by combining the environmental interference degree; Measure the variable data, perform normalization processing on the variable data, extract the variable change degree, and perform positioning according to the variable change degree to obtain the fault location result; The step of obtaining the environmental data of the area measurement points and evaluating the environmental interference degrees of different variables based on the environmental data is specifically as follows: Obtain the environmental data of the area measurement points in real time through a distributed sensor network, and the environmental data includes temperature W, humidity S, wind speed F, electromagnetic noise energy Z, and geomagnetic disturbance index C; Perform normalization processing and denoising processing on the collected environmental data to generate a standardized data set; Establish an environmental interference degree equation according to the standardized data set: Among them, DB is the electrical variable interference degree, and CB is the magnetic variable interference degree. is the temperature-humidity coupling function. is the wind speed effect function. is the geomagnetic disturbance coefficient. is the wind vibration magnetic circuit offset. is the influence function of temperature on magnetic permeability. , , , , , are proportionality coefficients, p is the label of different electrical variables, and q is the label of different magnetic variables. Calculate the environmental interference degrees of different variables according to the environmental interference degree equation.

2. The method for fault location of an overhead hybrid distribution line according to claim 1, wherein, The step of collecting the area data of the grid area points, evaluating the abnormal probability of the grid area points based on the area data, and selecting area measurement points according to the abnormal probability is specifically as follows: Collect the area data of the grid area points, and the area data includes environmental data and biological data; Statistically calculate the change differences of different environmental factors in the environmental data, and superimpose the change differences of all environmental factors to obtain the environmental change value; Obtain the line position of the overhead hybrid distribution line at the grid area points, and statistically calculate the presence rates of different organisms at the line position; Select the organisms with the presence rate reaching the preset presence rate threshold as the reflection organisms, and extract the reflection organism data from the biological data; Evaluate the change situation of the reflection organisms according to the reflection organism data to obtain the biological change value, combine the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line, and select the grid area point with the maximum abnormal probability as the area measurement point.

3. A fault location method for an overhead hybrid distribution line according to claim 2, characterized in that, The step of evaluating the change situation of the reflection organisms according to the reflection organism data to obtain the biological change value, combining the environmental change value to obtain the abnormal probability of the overhead hybrid distribution line is specifically as follows: Extract the change value of the habitat density of the reflection organisms according to the reflection organism data, and judge whether the change value of the habitat density reaches the preset density change threshold. If it reaches the preset density change threshold, use the change value of the habitat density as the biological abnormal value; If it does not reach the preset density change threshold, obtain the daily behavior characteristics of the reflection organisms, and extract the real-time behavior characteristics of the reflection organisms according to the reflection organism data; Compare the feature similarity between the daily behavior characteristics and the real-time behavior characteristics, and use the reciprocal of the feature similarity as the biological abnormal value; After normalizing all biological outliers, calculate the average value of all biological outliers reflecting the organism as the biological change value; Use the linear regression algorithm to fit the linear relationship between the environmental change value and the biological change value and the abnormal probability of the overhead hybrid distribution line, generate a change linear equation, and calculate the abnormal probability of the overhead hybrid distribution line according to the change linear equation.

4. A fault location method for an overhead hybrid distribution line according to claim 1, characterized in that, The steps of predicting the fault type through the monitoring data, obtaining the sensitivity of different variables to the fault type, collecting the measurement difficulty of the variables, and selecting the measured variables in combination with the environmental interference degree are specifically as follows: Obtain the historical fault data of the fault type, and extract the average change values of different variables during the fault from the historical fault data; Standardize the average change values of different variables to obtain the basic sensitivities of different variables; Obtain the line segment form of the overhead hybrid distribution line at the regional measurement point and record it as the measurement line segment. The line segment form is divided into an overhead line segment and a cable line segment; Obtain the resistivity and dynamic impedance fluctuation value of the overhead line segment, combine the basic sensitivity of the electrical variable to obtain the variable sensitivity of the electrical variable of the overhead line segment, and use the basic sensitivity of the magnetic variable as the variable sensitivity of the magnetic variable of the overhead line segment; Obtain the capacitance effect value of the cable line segment, and obtain the variable sensitivity of the electrical variable of the cable line segment according to the capacitance effect value and the basic sensitivity of the electrical variable; Obtain the shielding performance value of the cable line segment, and obtain the variable sensitivity of the magnetic variable of the cable line segment according to the shielding performance value and the basic sensitivity value of the magnetic variable; Obtain the variable sensitivities of the corresponding different variables according to the measurement line segment and record them as the line segment sensitivities, collect the measurement selectivities of the variables, calculate the measurement degrees of different variables in combination with the environmental interference degree, and select the variable with the largest measurement degree for measurement.

5. A fault location method for an overhead hybrid distribution line according to claim 4, characterized in that, The steps of collecting the measurement selectivity of the variables are specifically as follows: Obtain the measurement method of the variable, judge whether the measurement method of the variable is single. If it is single, judge whether the measured variable must be in physical contact with the measurement line segment to obtain a contact judgment result; Obtain the operation data for the measurement line segment to satisfy physical contact, and obtain the variable measurement difficulty according to the operation data and the contact judgment result; If the measurement method is not single, use the minimum variable measurement difficulty in the measurement method as the variable measurement difficulty of the variable; Evaluate the capture difficulty of the variable according to the historical fault data, obtain the known data, and evaluate the effectiveness of the variable according to the known data; Combine the variable measurement difficulty, capture difficulty and effectiveness to obtain the measurement selectivity of the variable.

6. The fault location method for an overhead hybrid distribution line according to claim 5, characterized in that, The steps of obtaining the operation data for the measurement line segment to satisfy physical contact and obtaining the variable measurement difficulty according to the operation data and the contact judgment result are specifically as follows: If it must be in physical contact with the measurement line segment, then count the operation duration, operation danger degree and operation influence degree according to the operation data; Evaluate the basic operation difficulty of the variable according to the operation data; Obtain the measurement steps of the variable, evaluate the basic measurement difficulty of the variable according to the measurement steps, and combine the basic operation difficulty to obtain the variable measurement difficulty; If it is not necessary to be in physical contact with the measurement line segment, record the basic measurement difficulty as the variable measurement difficulty.

7. A fault location method for an overhead hybrid distribution line according to claim 5, characterized in that, The steps of evaluating the capture difficulty of the variable according to the historical fault data are specifically as follows: Obtain the average duration of changes in different variables during a fault extracted from historical fault data; Estimate the degree of the fault based on the monitoring data, and estimate the dynamic range of variable fluctuations in combination with the line segment sensitivity of the variable; Determine whether the dynamic range is covered by a single measurement operation. If the dynamic range is covered by a single measurement operation, establish a correlation curve between the average duration and the capture difficulty, and find the corresponding capture difficulty according to the correlation curve; If the dynamic range is not covered by a single measurement operation, obtain the fluctuation frequency of variable fluctuations, and obtain the capture difficulty according to the fluctuation frequency.

8. A fault location method for an overhead hybrid distribution line according to claim 7, characterized in that, The step of, if the dynamic range is not covered by a single measurement operation, obtaining the fluctuation frequency of variable fluctuations and obtaining the capture difficulty according to the fluctuation frequency is specifically as follows: Obtain the measurement operation methods that cover the dynamic range, and obtain the coverage range of each measurement operation method; Record the variable fluctuation from the coverage range of one method to the coverage range of another method as one fluctuation, and count the fluctuation frequency of the variable; Statistically calculate the average existence duration of variable fluctuations in the coverage ranges of different methods, and obtain the capture difficulty in combination with the fluctuation frequency.

9. A fault location method for an overhead hybrid distribution line according to claim 5, characterized in that, The step of obtaining the known data and evaluating the validity of the variable based on the known data is specifically as follows: Obtain the known data of the overhead hybrid distribution line, and statistically calculate the calculation data by combining the variables and the known data; Screen the calculation data related to fault location as valid data, and find the variable sensitivity corresponding to the valid data; Sum all the valid data and the corresponding variable sensitivities to obtain the validity of the variable.

Images