Multi-dimensional microclimate-based ultra-high voltage transmission line abnormal line loss analysis method

Abstract

The application provides a multi-dimensional microclimate ultra-high voltage transmission line abnormal line loss analysis method, data mining is performed on historical data, abnormal line loss of a transmission line is identified and classified, discharge loss of the line under different climate conditions is considered, the ultra-high voltage transmission line is processed in sections in view of line loss difference caused by different section microclimate conditions, the line is divided into different line loss areas, a more accurate line loss analysis model is established for the line of different line loss areas, a line loss prediction model is established, line loss is predicted according to actual operation information of a power grid, abnormal line loss is identified and reasons are analyzed according to actual line loss and theoretical line loss, and an abnormal line loss treatment scheme is proposed based on the reasons.

Description

Technical Field

[0001] This invention relates to the field of power transmission technology, and more specifically, to a multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage transmission lines. Background Technology

[0002] The line loss rate of a power grid is a crucial economic and technical indicator for evaluating the power grid and the power industry. It accurately reflects the management level of my country's power sector in grid operation, and refined line loss calculation is beneficial for relevant departments to plan the power grid and manage its operational feasibility. The 1000 kV ultra-high voltage (UHV) transmission line corridor traverses complex terrains such as high mountains, valleys, and rivers, crossing the climatic north-south dividing line, with significant climate differences along the route. Inevitably, there is insufficient understanding and research into the micro-topography and micro-meteorology of these areas, failing to fully reflect the meteorological conditions of valleys, wind gaps, and rivers along the UHV line. This results in frequent occurrences of transmission line icing, wind vibration, and flashover after commissioning, significantly impacting line losses.

[0003] The method for calculating the reasonable range of distribution network line loss based on big data analysis, with publication number CN117117841A, uses energy efficiency guidelines and expert experience values ​​to establish a theoretical line loss benchmark to obtain a subjective range value; based on interval arithmetic theory and power flow algorithm, an objective theoretical line loss range is obtained; based on the big data analysis method of clustering decision tree, the influence of different distribution network conditions on the theoretical line loss value and the reference value of the line loss scale are obtained; feeders whose theoretical line loss calculation results are outside the range are identified as abnormal line loss points; and the primary and secondary causes of line loss are judged based on the correlation coefficient method.

[0004] However, the existing local power grid has formed a new operating situation similar to a small ring network under a large ring network. The line loss analysis is becoming increasingly complex. There is a lack of multi-dimensional line loss analysis methods based on UHV ring networks, and a lack of modeling and analysis of UHV ring network line losses under micro-meteorological conditions. It is impossible to obtain an accurate comprehensive line loss rate, and it is difficult to carry out targeted management of abnormal line losses. Summary of the Invention

[0005] This invention addresses the technical problems existing in the prior art by providing a multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage transmission lines, comprising:

[0006] Historical data of ultra-high voltage transmission lines are collected. The historical data includes multi-dimensional data, including line data, power generation data, meteorological data, topographic data, and daily line loss rate. The historical daily line loss rate is obtained by standard measuring instrument. The transmission line corridor is divided into different line loss zones based on topographic data and meteorological data.

[0007] For each line loss zone, based on the continuous daily line loss rate, the abnormal line loss type of the transmission line within the set time period is identified, and the correlation between the daily line loss rate of the identified abnormal line loss within the set time period and various influencing factors is analyzed.

[0008] Based on the correlation between the daily line loss rate and various influencing factors, sample data is screened. The sample data includes power consumption data, meteorological data, and daily line loss rate of transmission lines within a set time period that is identified as having abnormal line loss.

[0009] The line loss prediction model is trained based on the sample data;

[0010] Collect actual operating power data, meteorological data, and actual line loss rate of transmission line transformers. Input the actual operating power data and meteorological data into the trained line loss prediction model to obtain the theoretical line loss rate output by the line loss prediction model.

[0011] By comparing the actual line loss rate with the theoretical line loss rate, we can determine whether the actual line loss rate is abnormal and identify the factors that cause the abnormal line loss.

[0012] This invention provides a multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage (UHV) transmission lines. It utilizes data mining of historical data to identify and classify abnormal line losses, considers discharge losses under different climatic conditions, and segments the UHV transmission line into different loss zones based on the differences in line losses caused by micro-meteorological conditions in different sections. More accurate line loss analysis models are established for lines in different loss zones, and line loss prediction models are also built. Based on actual grid operation information, line losses are predicted. Based on actual and theoretical line losses, abnormal line losses are identified and their causes analyzed, and mitigation solutions for abnormal line losses are proposed based on these causes. Attached Figure Description

[0013] Figure 1 A flowchart of a multi-dimensional micrometeorological method for analyzing abnormal line losses in ultra-high voltage transmission lines provided by this invention; Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. In addition, the technical features of the various embodiments or individual embodiments provided by the present invention can be arbitrarily combined with each other to form feasible technical solutions. Such combinations are not constrained by the order of steps and / or structural composition patterns, but must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0015] Figure 1 A flowchart of a multi-dimensional micrometeorological method for analyzing abnormal line losses in ultra-high voltage transmission lines, as provided by this invention, is shown below. Figure 1 As shown, the method includes:

[0016] Step 1: Collect historical data of UHV transmission lines. The historical data includes multi-dimensional data, including line data, power generation data, meteorological data, terrain data, and daily line loss rate; and divide the transmission line corridor into different line loss zones based on the terrain data and meteorological data.

[0017] Understandably, historical data on UHV transmission lines is collected. This historical data includes multidimensional data, such as line data, power generation data, meteorological data, topographic data, and daily line loss rates. It should be noted that collecting this historical data is to provide training data for subsequent line loss prediction models. The daily line loss rate can be directly measured using a measuring instrument. When collecting historical data, the daily line loss rate was measured using a standard measuring instrument to ensure accuracy.

[0018] The data includes conductor structure, wire diameter, number of splits, split spacing, phase distance, and height above ground; electrical data includes voltage, current, load rate, active power, and reactive power; meteorological data includes ambient temperature, humidity, wind speed, wind direction, and rainfall; and topographic data includes altitude and latitude / longitude.

[0019] Based on the collected data on transmission lines, the micro-meteorological influencing factors of UHV transmission lines were sorted out, and the transmission line corridors were divided into general line loss areas and micro-meteorological line loss areas according to the discharge characteristics of UHV.

[0020] General line loss areas refer to the line losses experienced by long-distance, ultra-high-voltage transmission lines under most operating conditions, with single influencing factors and similar environmental factors. Micro-meteorological line loss areas refer to a small section of a transmission line, or even just one or two towers, where local weather conditions, particularly influenced by specific terrain, intensify certain climatic factors, exceeding the designed icing and wind conditions for the region. Under these circumstances, discharge phenomena are more pronounced than on normal lines, resulting in increased losses.

[0021] Micro-meteorological and micro-topographical regions include: micro-topographical regions such as mountains, passes, and rivers, as well as micro-meteorological regions such as icing, galloping, micro-wind vibration, and insulator contamination.

[0022] Line losses include resistance loss, corona loss, and overhead ground wire loss. Abnormal line losses originate from abnormal resistance losses of UHV long-distance transmission lines under different operating modes, as well as abnormal corona losses generated in a local micro-meteorological area.

[0023] Step 2: For each line loss zone, based on the continuous daily line loss rate, identify the abnormal line loss type of the transmission line within the set time period, and analyze the correlation between the daily line loss rate of the identified abnormal line loss within the set time period and various influencing factors.

[0024] Understandably, based on the composition of line losses under 1000 kV UHV micrometeorological conditions, the quartile algorithm is used to identify abnormal line losses. Combined with grey relational analysis and multi-dimensional data analysis, the correlation between line losses and micrometeorological conditions is analyzed to identify micrometeorological factors that characterize the causes of line losses and related line information.

[0025] Among them, abnormal line loss types include long-term negative line loss, long-term high line loss, long-term positive and negative line loss, and sudden high loss, which are analyzed in detail as follows:

[0026] 1) Long-term negative line loss: The judgment condition is that the daily line loss rate is <0% for 11 or more days within 15 consecutive days;

[0027] Influencing factors include abnormal data acquisition and abnormal technical line loss (dual-circuit operation mode, low load rate, and line electric field distortion).

[0028] 2) Long-term high line loss: The judgment condition is that the daily line loss rate is >7% for 11 or more days within 15 consecutive days;

[0029] Influencing factors include metering issues, data collection issues, electricity theft, and high technical line losses (high load rate, long power supply radius, and high micro-meteorological line losses).

[0030] 3) Sudden High Loss: The judgment criterion is that, based on the box plot quartile method, the data from the detection day is greater than (Q3 + 1.5Q) in the 15-day sample value. R );

[0031] The influencing factors include electricity theft, data collection issues, and micro-meteorological line loss (localized strong winds).

[0032] 4) Long-term positive and negative line loss: The judgment condition is that the daily line loss rate is <0 for 6 or more days within 15 consecutive days, and the daily line loss rate is >=0 for 6 or more days.

[0033] The influencing factors are clock problems or abnormal household-transfer relationship issues.

[0034] Specifically, the use of the box plot quartile method to determine abnormal line loss is as follows:

[0035] Using (Q) R This method detects outliers and provides a standard for identifying them: outliers are typically defined as values ​​less than Q1-1.5Q. R or greater than Q3 + 1.5Q R .

[0036] Q1: Lower quartile, indicating that one-quarter of all observed values ​​are smaller than it; Q3: Upper quartile, indicating that one-quarter of all observed values ​​are larger than it; Q R The interquartile range (IQR) is the difference between the upper quartile (Q3) and the lower quartile (Q1), encompassing half of all observations. The minimum estimated value is: Q1 - k * Q. R The maximum estimated value is: Q3 + k * Q R k = 1.5 (moderately abnormal), k = 3 (severely abnormal)

[0037] The threshold for normal data is k times the quartile; data outside this threshold is considered outlier. Here, k is set to 1.5, meaning the daily line loss rate is greater than Q3 + 1.5Q. R At that time, it was considered that the line loss rate suddenly increased, and the line loss rate on that day was less than Q1-1.5Q. R At that time, it was believed that the line loss rate suddenly decreased.

[0038] Using the methods described above, different types of abnormal line losses were identified, and the influencing factors for each type of abnormal line loss were analyzed.

[0039] The grey relational analysis method is used below to analyze the correlation between abnormal line losses and various influencing factors. The stronger the correlation, the greater the impact of the influencing factor on the line loss; the weaker the correlation, the smaller the impact. The influencing factors here mainly refer to meteorological factors.

[0040] In this context, the daily line loss rate within a set time period that is identified as abnormal line loss is recorded as a reference sequence, and the various influencing factors corresponding to the daily line loss rate are recorded as a comparison sequence. The reference sequence and the comparison sequence are represented in matrix form.

[0041] The reference sequence is the line loss {X0(j)} at each time point, where j = 1, 2, ..., N, and N is the number of time points. The comparison sequence is the time series {X0(j)} composed of the effective influencing factors of line loss. i (j)}, where i = 1, 2, ..., M, i.e., M influencing factors, the reference sequence and comparison sequence are represented in matrix form, forming matrix X:

[0042]

[0043] In the formula, X 0j As the reference sequence, i.e., the linear loss sequence, X ij The comparison sequence is the time series composed of various factors affecting line loss, i = 1, 2, ..., M; j = 1, 2, ..., N.

[0044] To transform the data in matrix X, the original data needs to be processed to become dimensionless data of roughly the same order of magnitude. This is done by using interval value transformation, represented as follows:

[0045]

[0046] By X ij The new matrix X' is formed.

[0047] Find the time series difference between the comparison sequence and the reference sequence in matrix X′:

[0048] Δ i (j)=|X0′(j)-X i ′(j)|;

[0049] Find the minimum difference between the two levels. i min j Δ i (j) and the maximum difference max i max j Δ i (j);

[0050] Calculate the correlation coefficient L ij :

[0051]

[0052] L ij Let denot represent the correlation coefficient between the j-th line loss and the i-th influencing factor, and ρ be the resolution coefficient, which takes values ​​in (0,1). It reflects the degree of indirect influence of each factor in the system on the correlation degree.

[0053] Step 3: Based on the correlation between the daily line loss rate and various influencing factors, filter sample data. The sample data includes power generation data, meteorological data, and daily line loss rate of transmission lines within a set time period that are identified as having abnormal line loss.

[0054] Understandably, the sample data is filtered based on the correlation coefficient between line loss and various influencing factors calculated in step 2. Specifically, if the correlation coefficient between the daily line loss rate and the influencing factors is greater than a set threshold, the daily line loss rate and the influencing factors are retained, and the power consumption data of the transmission line on that day is obtained. The daily line loss rate, power consumption data and influencing factors (meteorological data) are combined to form the sample data.

[0055] Step 4: Train the line loss prediction model based on the sample data.

[0056] It should be noted that the line loss prediction model in this invention, and the line loss analysis, involves segmenting the transmission line, establishing a distributed micro-meteorological line loss analysis model based on the grid structure, transmission power, line corridor, and micro-topography and micro-meteorological conditions, and using an improved RBF neural network algorithm to predict line loss, thereby obtaining the line loss rate of the transmission line under micro-meteorological conditions.

[0057] The path traversed by ultra-high voltage long-distance transmission lines is segmented, and the impact of different variables such as temperature, humidity, air density, wind speed, and rainfall on line loss is studied in the common line loss area.

[0058] The microclimate zone is divided into icing zone, galloping zone, light wind vibration zone, and insulator pollution zone, with the influencing factor being the localized excessively strong climatic factors.

[0059] An improved RBF neural network was used to construct a line loss prediction model. The grey relational method was used to improve the training process of the RBF neural network, avoiding the problem of parameters (center vector, extension constant) getting trapped in local minima, and optimizing the test accuracy and generalization ability.

[0060] Considering the layout of the UHV ring network, line loss prediction was performed for the special power flow of the small ring network under the large ring network, and the line loss prediction model was improved. The training process of the line loss prediction model is as follows:

[0061] 1) Based on historical data of the line, the input and output samples of the RBF neural network algorithm are formed. The input samples are denoted as S = {(S1, S2, ..., S...} m Let Q be the output sample and}.

[0062] 2) Select a corresponding number of samples from the input samples as cluster centers based on the number m of neurons in the hidden layer of the RBF neural network. The clustering algorithm will divide the input samples into several different clusters, and the center of each cluster is a center vector of the RBF function.

[0063] 3) Initialize the weights of the RBF neural network. The input samples pass through the RBF neural network, performing operations from the input layer to the output layer. The error between the output node and Q is calculated, expressed by the formula:

[0064]

[0065] In the formula, P represents the number of output nodes; Represents the actual value; Q p This represents the computation result of the RBF neural network.

[0066] 4) Calculate the fitness values ​​of the RBF neural network weights. Based on the fitness evaluation results, select individuals with higher fitness to enter the next generation of the population, and perform crossover and mutation operations on the selected individuals to generate new individuals.

[0067] The formula for calculating the fitness value (FIT) is:

[0068]

[0069] 5) Repeat the above process until the preset number of iterations is reached.

[0070] 6) Select the individual with the highest fitness from the population optimized by the genetic algorithm, and decode to obtain the optimal RBF neural network parameters.

[0071] A grey relational analysis method was used to screen climate characteristic parameters and combine them with electrical and line characteristic parameters to form an influence characteristic system for UHV transmission line losses, which served as the input. Controlled power flow (RBF) was used to calculate line losses, which served as the output. The RBF prediction model was trained using these inputs and outputs, resulting in a theoretical line loss with small error and high reliability.

[0072] Step 5: Collect actual operating power data, meteorological data, and actual line loss rate of the transmission line transformer. Input the actual operating power data and meteorological data into the trained line loss prediction model to obtain the theoretical line loss rate output by the line loss prediction model.

[0073] Understandably, step 4 trains the line loss prediction model. This step uses the trained model to predict line losses in transmission lines and identify anomalies. Specifically, it collects actual operating power data and meteorological data from the transmission line transformers and measures the actual line loss rate using a measuring instrument. The actual operating power data and meteorological data are then input into the trained line loss prediction model to obtain the theoretical line loss rate output by the model.

[0074] Step 6: Compare the actual line loss rate with the theoretical line loss rate to determine whether the actual line loss rate is abnormal and identify the influencing factors of the abnormal line loss.

[0075] Understandably, the theoretical line loss rate predicted by the line loss prediction model is generally considered to be correct. The actual line loss rate measured by the measuring instrument is compared with the theoretical line loss rate. If the difference between the actual line loss rate and the theoretical line loss rate is within a certain range, it indicates that the actual line loss rate is normal. If the difference exceeds a certain range, it indicates that the actual line loss rate is abnormal. At this time, the cause of the abnormal line loss should be investigated. It may be due to the inaccuracy of the measuring instrument or other factors.

[0076] Based on the results of the abnormal line loss rate analysis, specific governance suggestions are proposed for the influencing factors of statistical line loss, including load level governance, electricity metering device error governance, electricity metering device governance, power grid structure adjustment, and technical line loss governance.

[0077] Load level management: Adjust the current transformer ratio appropriately to ensure that the metering equipment operates under rated conditions.

[0078] Electricity metering device errors: Metering device errors include system truncation errors, time errors, etc. For system truncation errors, the number of decimal places set after the power acquisition system is set to 5, and the line loss statistical period is lengthened; for clock errors, the time management of the electricity meter is strengthened to ensure that the electricity meter clock is synchronized with the standard clock.

[0079] Management of electricity metering devices: Regular testing is conducted, and devices with substandard errors are replaced promptly. Based on the technical regulations and standards for the management of electricity metering instruments, strict control is exercised over the selection, procurement, inspection, and acceptance of electricity metering devices.

[0080] Power grid structure adjustment: Reasonable planning of the power grid ring network structure, selection of the optimal power source point, adjustment of line operation mode, optimization of power flow direction, thereby controlling line loss.

[0081] Technical line loss management: For lines with high load rates, special line upgrades are carried out to enhance power supply stability. Based on data mining and line loss analysis results, power flow control is implemented in micro-meteorological areas to reduce partial discharge losses.

[0082] The present invention provides a multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage transmission lines, which has the following beneficial effects:

[0083] (1) A multi-dimensional micro-meteorological line loss analysis method has been developed. This method is based on historical line loss data and big data analysis, and takes into account the differences in power generation loss caused by different micro-meteorological conditions under actual operating conditions. The line loss analysis results are practical and reliable.

[0084] (2) The changes in line loss under micro-meteorological conditions were taken into consideration. Based on the traditional line, the influence of capacitance and conductivity parameters on line discharge was considered.

[0085] (3) Combining the grey relational analysis method and the improved RBF neural network for multi-dimensional data analysis, the correlation between line loss and micrometeorological conditions is analyzed to identify micrometeorological factors characterizing the causes of line loss and related line information.

[0086] (4) Based on the line operation under complex micro-meteorological conditions, a targeted analysis was conducted, taking into account the general line loss area and the micro-meteorological line loss area. The micro-meteorological line loss area is further divided into icing area, galloping area, micro-wind vibration area and insulator pollution area, which ensures the accuracy of line loss classification.

[0087] (5) Multidimensional data includes electrical data such as voltage, current, active power, and reactive power, as well as weather data such as temperature, humidity, and rainfall, line data such as line parameters, topology, and electrical corridors, and micro-meteorological data such as line icing, pollution, and galloping.

[0088] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0089] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0090] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0091] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0092] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0093] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0094] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage transmission lines, characterized in that, include: Historical data of ultra-high voltage transmission lines are collected. The historical data includes multi-dimensional data, including line data, power generation data, meteorological data, topographic data, and daily line loss rate. The historical daily line loss rate is obtained by standard measuring instrument. The transmission line corridor is divided into different line loss zones based on topographic data and meteorological data. For each line loss zone, based on the continuous daily line loss rate, the abnormal line loss type of the transmission line within a set time period is identified, and the correlation between the daily line loss rate of the set time period identified as abnormal line loss and various influencing factors is analyzed. Based on the correlation between the daily line loss rate and various influencing factors, sample data is screened. The sample data includes power consumption data, meteorological data, and daily line loss rate of transmission lines within a set time period that are identified as having abnormal line loss. The line loss prediction model is trained based on the sample data; Collect actual operating power data, meteorological data, and actual line loss rate of transmission line transformers. Input the actual operating power data and meteorological data into the trained line loss prediction model to obtain the theoretical line loss rate output by the line loss prediction model. By comparing the actual line loss rate with the theoretical line loss rate, we can determine whether the actual line loss rate is abnormal and identify the factors that cause the abnormal line loss.

2. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 1, characterized in that, The line data includes conductor structure, wire diameter, number of splits, split spacing, phase distance, and height above ground; the electrical data includes voltage, current, load rate, active power, and reactive power; the meteorological data includes ambient temperature, humidity, wind speed, wind direction, and rainfall; and the terrain data includes altitude and latitude and longitude.

3. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 1, characterized in that, The abnormal line loss types include long-term negative line loss, long-term high line loss, long-term positive and negative line loss, and sudden high loss. For each line loss zone, based on continuous daily line loss rates, the identification of abnormal line loss types for transmission lines within a set time period includes: If the daily line loss rate is less than 0% for 11 or more days within 15 consecutive days, then the output line has a long-term negative line loss within 15 consecutive days. The influencing factors include abnormal acquisition and / or abnormal technical line loss, which includes dual-circuit operation mode, low load rate and / or line electric field distortion. If the daily line loss rate is >7% for 11 or more days within a consecutive 15 days, then the output line is considered to have long-term high line loss for 15 consecutive days. The influencing factors include metering problems, data collection problems, electricity theft, and / or high technical line loss. High technical line loss includes high load rate, long power supply radius, and / or high micro-meteorological line loss. The box plot quartile method was used to determine whether the data from the detection day was greater than (Q3 + 1.5Q) in the 15-day sample. R ), where Q3 is the upper quartile, Q R If the interval is quartile, then the output line will experience a sudden high loss within 15 consecutive days. The influencing factors are electricity theft, data collection issues, and / or micro-meteorological line loss issues, including local strong winds. If the daily line loss rate is less than 0 for 6 or more days within a consecutive 15-day period, and the daily line loss rate is greater than or equal to 0 for 6 or more days, then the output line will have long-term positive and negative line loss for 15 consecutive days. The influencing factors are clock problems and / or abnormal user-transformer relationship problems.

4. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 3, characterized in that, Identifying abnormal line loss using the box plot quartile method includes: Using Q R Outlier detection was performed, and anomaly identification criteria were set: outliers were defined as values ​​less than Q1-1.5Q. R or greater than Q3 + 1.5Q R ; Where Q1 is the lower quartile, representing one-quarter of all observed values ​​that are smaller than it; Q3 is the upper quartile, representing one-quarter of all observed values ​​that are larger than it; Q... R The interquartile range is the difference between the upper quartile Q3 and the lower quartile Q1, encompassing half of all observations; the minimum estimate is: Q1-k. Q R The maximum estimated value is: Q3+k Q R； The daily line loss rate is greater than Q3+kQ. R At that time, it was considered that the line loss rate suddenly increased, and the line loss rate on that day was less than Q1-kQ. R At that time, it was assumed that the line loss rate had suddenly decreased; The value of k varies depending on the degree of abnormality. When it is necessary to judge moderate outliers, k = 1.5, and when it is necessary to judge severe outliers, k = 3.

5. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 1, characterized in that, The analysis identifies the correlation between the daily line loss rate and various influencing factors within a defined time period for those identified as having abnormal line losses, including: The daily line loss rate within a set time period identified as abnormal line loss is recorded as a reference sequence, and the influencing factors corresponding to the daily line loss rate are recorded as a comparison sequence. The reference sequence and the comparison sequence are represented in matrix form. The data in matrix form is processed using interval value transformation to form a new matrix; Calculate the time series difference between the reference sequence and the comparison sequence in the new matrix; Based on the time series difference between the reference sequence and the comparison sequence, calculate the minimum and maximum differences at two levels; Based on the minimum and maximum differences at two levels, the correlation between the daily line loss rate and various influencing factors is calculated.

6. The multi-dimensional micro-meteorological analysis method for abnormal line losses in ultra-high voltage transmission lines according to claim 5, characterized in that, The reference sequence is the line loss {X0(j)} at each time point, where j = 1, 2, ..., N, and N is the number of time points. The comparison sequence is a time series {X0(j)} composed of the effective influencing factors of line loss. i (j)}, where i = 1, 2, ..., M, i.e., M influencing factors, the reference sequence and comparison sequence are represented in matrix form to form matrix X: ； In the formula, X 0j As the reference sequence, i.e., the linear loss sequence, X ij The comparison sequence is the time series composed of various factors affecting line loss, i=1,2,…,M; j = 1, 2, ..., N; The process of transforming data into a matrix using interval-valued methods to form a new matrix includes: ； The time series difference between the reference sequence and the comparison sequence is expressed as: ； The minimum difference between the two levels is expressed as: The maximum difference is expressed as ; The calculation of the correlation coefficient between the daily line loss rate and various influencing factors includes: ； Wherein, ρ is the resolution coefficient, which takes values ​​in (0,1), and it reflects the degree of indirect influence of each influencing factor on the line loss rate.

7. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 1 or 6, characterized in that, The process of filtering sample data based on the correlation between the daily line loss rate and various influencing factors includes: If the correlation coefficient between the daily line loss rate and the influencing factors is greater than a set threshold, the daily line loss rate and the influencing factors are retained, and the power consumption data of the transmission line on that day is obtained. The daily line loss rate, power consumption data and influencing factors are then combined to form sample data.

8. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in UHV transmission lines according to claim 7, characterized in that, The line loss prediction model is an RBF neural network. The training process of the RBF neural network is improved using the grey relational analysis method. The training process of the RBF neural network includes: Based on historical data of the railway line, the input and output samples of the RBF neural network algorithm are composed. The input samples are denoted as S={(S1, S2, ..., S...}. m } and the output sample are denoted as ; Based on the number m of neurons in the hidden layer of the RBF neural network, a corresponding number of samples are selected from the input samples as cluster centers. The clustering algorithm is used to divide the input samples into different clusters, and the center of each cluster is a center vector of the RBF function. Initialize the weights of the RBF neural network, pass the input samples through the RBF neural network, perform the RBF neural network operations from the input layer to the output layer, and calculate the output node results. The error is expressed by the formula: ； In the formula, This indicates the error, and P represents the number of output nodes. Represents the actual value. This represents the computation result of the RBF neural network; The fitness values ​​of the RBF neural network weights are calculated. Based on the fitness evaluation results, individuals with high fitness are selected to enter the next generation of the population. Crossover and mutation operations are then performed on the selected individuals to generate new individuals. fitness The calculation formula is: ； Repeat the iterative process until the preset maximum number of iterations is reached; The individual with the highest fitness is selected from the population optimized by the genetic algorithm, and the optimal RBF neural network parameters are decoded to obtain the trained line loss prediction model.

9. The multi-dimensional micro-meteorological method for analyzing abnormal line losses in ultra-high voltage transmission lines according to claim 1, characterized in that, The process of comparing the actual line loss rate with the theoretical line loss rate to determine whether the actual line loss rate is abnormal and to identify the influencing factors of abnormal line loss includes: If the actual line loss rate is abnormal, specific governance recommendations will be made for the influencing factors of statistical line loss based on the results of the line loss rate anomaly analysis, including load level governance, electricity metering device error governance, electricity metering device governance, power grid structure adjustment and technical line loss governance.

Images