CN115905904A

CN115905904A - Line loss abnormity evaluation method and device for power distribution network feeder line

Info

Publication number: CN115905904A
Application number: CN202211507531.0A
Authority: CN
Inventors: 江泽涛; 李固; 李健; 招景明; 马喆非; 党三磊; 蔡永智; 赵闻; 姚智聪; 彭龙
Original assignee: Guangdong Power Grid Co Ltd; Measurement Center of Guangdong Power Grid Co Ltd
Current assignee: Guangdong Power Grid Co Ltd; Measurement Center of Guangdong Power Grid Co Ltd
Priority date: 2022-11-28
Filing date: 2022-11-28
Publication date: 2023-04-04

Abstract

The invention discloses a method and a device for evaluating line loss abnormity of a feeder line of a power distribution network, wherein a power distribution network line loss characteristic index system containing a distributed power supply is constructed, and collected power distribution network line loss data and line loss characteristic index data of a line area are used for training an LSTM neural network model to obtain an optimal power distribution network line loss prediction model, so that line loss prediction of line loss characteristic index data to be predicted is carried out based on the optimal power distribution network line loss prediction model, and a line loss reasonable confidence interval of the power distribution network line loss is generated based on the obtained first power distribution network line loss prediction data; clustering the first power distribution network line loss prediction data based on the line loss reasonable confidence interval and an improved fuzzy C-means algorithm to obtain a suspected abnormal feeder, and evaluating the abnormal degree of the power distribution network line loss by calculating an abnormal coefficient of the suspected abnormal feeder. Compared with the prior art, the technical scheme of the invention can improve the accuracy of the power distribution network line loss abnormity evaluation by establishing the abnormity coefficient to evaluate the line loss abnormity degree.

Description

Line loss abnormity evaluation method and device for power distribution network feeder line

Technical Field

The invention relates to the technical field of power distribution network line loss prediction, in particular to a method and a device for evaluating line loss abnormity of a power distribution network feeder line.

Background

With the proposal of the national 'double-carbon' target, the renewable energy source substitution action is implemented, and the construction of a novel power system taking new energy sources as main bodies is the important strategic deployment of the country; distributed power supplies such as wind power generation, photovoltaic power generation and the like are connected to a power distribution network in large quantity, so that the operation parameters of the power grid are greatly influenced; on one hand, the accessed distributed power supply can affect the distribution network tide and directly affect the line loss; on the other hand, the instability of the output power can bring a series of problems such as harmonic waves, voltage fluctuation and the like, and the quality of the power supply voltage of the power distribution network is affected. Disturbance brought by the distributed power supply interacts with original disturbance factors of the power distribution network, so that a power quality disturbance mechanism is more complex, and the difficulty of loss analysis of the power distribution network is increased. Therefore, the influence of distributed power supply access on the loss of the power distribution network is analyzed by combining the actual development condition of the power distribution network, the abnormal line loss is accurately diagnosed, and the method has important significance for effectively developing energy-saving and loss-reducing work and improving the economic benefit of power enterprises.

For the line loss abnormity diagnosis of the power distribution network, a line loss analysis model based on an intelligent ammeter is researched at present, and the line loss abnormity identification of the power distribution network is realized through a clustering algorithm; according to the line loss rate of the power distribution station area, the line loss abnormity is distinguished based on a k-means clustering algorithm, and whether the line loss abnormity exists in the power distribution station area can be judged. Therefore, the method has very important research significance for analyzing the line loss abnormity by using the data mining technology. Although there has been relatively comprehensive research on analysis of line loss influence factors, in a novel power system environment, the research does not fully consider the influence of distributed power supply access on line loss characteristics, and particularly, the current research on the time-space distribution characteristics of the power system is still lacking, so that the line loss cannot be effectively analyzed in time-space, and the situation that the distributed power supply in an actual power distribution network is flexible and changeable is difficult to adapt to. Moreover, the current research is mainly limited to the line loss theoretical calculation under the background of the traditional power system, the proportion and the type of new energy are not sufficiently considered, and a line loss reasonable interval with reference significance is difficult to form.

Most of the existing line loss abnormal diagnosis at home and abroad analyzes the line loss abnormal data from the perspective of data mining, so that the judgment efficiency is improved to a certain extent, but the accuracy of abnormal line loss diagnosis may be influenced due to incomplete characteristic indexes. Moreover, in many diagnosis schemes, only the traditional power distribution network is considered, the influence of the dynamic change of the distributed power supply on line loss characteristics is not considered, secondary analysis on negative high-loss lines is not available, the method is difficult to apply to a novel power system with new energy as a main body, and accurate diagnosis of abnormal line loss of the power distribution network cannot be realized.

Disclosure of Invention

The technical problem to be solved by the invention is as follows: the method and the device for evaluating the line loss abnormity of the power distribution network feeder line can improve the accuracy of evaluating the line loss abnormity of the power distribution network.

In order to solve the technical problem, the invention provides a line loss abnormity evaluation method of a power distribution network feeder line, which comprises the following steps:

acquiring and constructing a power distribution network line loss characteristic index system containing a distributed power supply based on line loss characteristic index data, wherein the line loss characteristic index data comprises weather factors, equipment grid-connected parameters and distribution network operation data;

collecting all power distribution network line loss data of a selected line area within preset time, collecting all line loss characteristic index data of the line area within the preset time based on a power distribution network line loss characteristic index system, respectively carrying out data processing on all power distribution network line loss data and all line loss characteristic index data to obtain standard power distribution network line loss data and standard line loss characteristic index data, and dividing the standard power distribution network line loss data and the standard line loss characteristic index data into a training data set and a testing data set;

constructing an LSTM neural network model based on an LSTM algorithm, performing model training on the LSTM neural network model according to the training data set and the test data set, and determining an optimal power distribution network line loss prediction model;

acquiring characteristic index data of line loss to be predicted of different distributed power types, inputting the characteristic index data of the line loss to be predicted into the optimal power distribution network line loss prediction model, outputting and calculating line loss reasonable confidence intervals of the power distribution network line loss of the different distributed power types according to first power distribution network line loss prediction data corresponding to the different distributed power types;

the method comprises the steps of performing primary clustering on first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, performing secondary clustering on the primary clustering result based on a line loss reasonable confidence interval to obtain a suspected abnormal feeder, calculating an abnormal coefficient of the suspected abnormal feeder, and evaluating and obtaining the power distribution network line loss abnormal degree of the suspected abnormal feeder according to the abnormal coefficient.

In a possible implementation manner, all power distribution network line loss data of the selected line region within a preset time are collected, and all line loss characteristic index data of the line region within the preset time are collected based on the power distribution network line loss characteristic index system, and the method specifically includes:

the method comprises the steps of obtaining the permeability of the distributed power supply of each line area in the power distribution network, selecting the line area of which the permeability is larger than a preset permeability threshold of the distributed power supply, and collecting line loss data of the power distribution network of the line area within one year;

and collecting line loss characteristic index data of the line region based on the power distribution network line loss characteristic index system.

In a possible implementation manner, the data processing is performed on the line loss data of all the power distribution networks and the characteristic index data of all the line losses, so as to obtain the line loss data of the standard power distribution network and the characteristic index data of the standard line loss, and the method specifically includes:

removing negative values of the power distribution network line loss data in all the power distribution network line loss data, and removing the power distribution network line loss data with the line loss rate of more than 25% in all the power distribution network line loss data to obtain first power distribution network line loss data;

judging whether data missing values exist in all the first power distribution network line loss data or all the line loss characteristic index data or not, if so, judging whether all the first power distribution network line loss data or all the line loss characteristic index data are in accordance with uniform distribution or not, if so, acquiring the mean value of all the first power distribution network line loss data or all the line loss characteristic index data and filling the mean value into the data missing values, otherwise, acquiring the median of all the first power distribution network line loss data or all the line loss characteristic index data and filling the median into the data missing values to obtain standard power distribution network line loss data and first line loss characteristic index data;

and carrying out normalization processing on the first line loss characteristic index data to obtain standard line loss characteristic index data.

In a possible implementation manner, an LSTM neural network model is constructed based on an LSTM algorithm, model training is performed on the LSTM neural network model according to the training data set and the test data set, and an optimal power distribution network line loss prediction model is determined, which specifically includes:

constructing an LSTM neural network model based on an LSTM algorithm, inputting the training data set into the LSTM neural network model for model training to obtain an initial power distribution network line loss prediction model, and inputting the test data set into the initial power distribution network line loss prediction model to obtain power distribution network line loss prediction data corresponding to the test data set;

and comparing the power distribution network line loss prediction data with the standard power distribution network line loss data to determine an optimal power distribution network line loss prediction model.

In a possible implementation mode, the line loss characteristic index data includes weather factors, equipment grid-connected parameters and distribution network operation data, wherein the weather factors include average temperature, average wind speed, illumination duration, illumination intensity and rainfall, the equipment grid-connected parameters include grid-connected capacity, grid-connected position, grid-connected operation mode and grid-connected electric quantity, and the distribution network operation data includes data date, power type, power supply radius, line load rate and power supply quantity.

In a possible implementation manner, obtaining to-be-predicted line loss characteristic index data of different distributed power supply types, and inputting the to-be-predicted line loss characteristic index data to the optimal power distribution network line loss prediction model specifically includes:

acquiring first to-be-predicted line loss characteristic index data of a distributed power supply type for wind power generation, wherein weather factors in the first to-be-predicted line loss characteristic index data comprise average temperature and average wind speed;

acquiring second line loss characteristic index data to be predicted, wherein the type of the distributed power supply is photovoltaic power generation, and weather factors in the second line loss characteristic index data to be predicted comprise illumination duration and illumination intensity;

acquiring third line loss characteristic index data to be predicted, wherein the distributed power supply is in a hydroelectric power generation type, and weather factors in the third line loss characteristic index data to be predicted comprise average temperature and rainfall;

and inputting the first line loss characteristic index data to be predicted, the second line loss characteristic index data to be predicted and the third line loss characteristic index data to be predicted into the optimal power distribution network line loss prediction model.

In a possible implementation manner, calculating an abnormal coefficient of the suspected abnormal feeder line, and evaluating and obtaining a distribution network line loss abnormal degree of the suspected abnormal feeder line according to the abnormal coefficient specifically includes:

taking a clustering center of the category of the suspected abnormal feeder line loss as a reference feeder line, acquiring suspected abnormal feeder line loss prediction data and suspected abnormal feeder line loss characteristic index data of the suspected abnormal feeder line, and acquiring reference feeder line loss prediction data and reference feeder line loss characteristic index data of the reference feeder line;

calculating the difference value between the suspected abnormal feeder line loss prediction data corresponding to the first preset time and the reference feeder line loss prediction data, and calculating the average value of the difference values;

calculating the deviation between the suspected abnormal feeder line loss characteristic index data and the reference feeder line loss characteristic index data, and calculating the deviation average value corresponding to each line loss characteristic index based on the number of the line loss characteristic indexes;

calculating and obtaining an abnormal coefficient of the suspected abnormal feeder line according to the difference average value and the deviation average value;

and comparing the abnormal coefficient with a preset abnormal coefficient threshold, if the abnormal coefficient is greater than or equal to the preset abnormal coefficient threshold, considering that the distribution network line loss abnormal degree of the suspected abnormal feeder line is higher, and if the abnormal coefficient is less than the preset abnormal coefficient threshold, considering that the distribution network line loss abnormal degree of the suspected abnormal feeder line is lower.

The invention also provides a device for evaluating the line loss abnormity of the feeder line of the power distribution network, which comprises: the system comprises a power distribution network line loss characteristic index system building module, a power distribution network data processing module, a power distribution network line loss prediction model building module, a line loss reasonable confidence interval calculating module and a power distribution network line loss abnormal degree evaluating module;

the power distribution network line loss characteristic index system construction module is used for acquiring and constructing a power distribution network line loss characteristic index system containing a distributed power supply based on line loss characteristic index data, wherein the line loss characteristic index data comprises weather factors, equipment grid-connected parameters and distribution network operation data;

the power distribution network data processing module is used for collecting all power distribution network line loss data of the selected line area within preset time, collecting all line loss characteristic index data of the line area within the preset time based on the power distribution network line loss characteristic index system, respectively carrying out data processing on all power distribution network line loss data and all line loss characteristic index data to obtain standard power distribution network line loss data and standard line loss characteristic index data, and dividing the standard power distribution network line loss data and the standard line loss characteristic index data into a training data set and a testing data set;

the power distribution network line loss prediction model construction module is used for constructing an LSTM neural network model based on an LSTM algorithm, performing model training on the LSTM neural network model according to the training data set and the test data set, and determining an optimal power distribution network line loss prediction model;

the line loss reasonable confidence interval calculation module is used for acquiring line loss characteristic index data to be predicted of different distributed power types, inputting the line loss characteristic index data to be predicted into the optimal power distribution network line loss prediction model, outputting and calculating line loss reasonable confidence intervals of the power distribution network line losses of different distributed power types according to first power distribution network line loss prediction data corresponding to different distributed power types;

the power distribution network line loss abnormal degree evaluation module is used for performing primary clustering on first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, performing secondary clustering on the primary clustering result based on the line loss reasonable confidence interval to obtain a suspected abnormal feeder line, calculating an abnormal coefficient of the suspected abnormal feeder line, and evaluating and obtaining the power distribution network line loss abnormal degree of the suspected abnormal feeder line according to the abnormal coefficient.

In a possible implementation mode, the power distribution network data processing module is configured to collect all power distribution network line loss data of the selected line region within a preset time, and collect all line loss characteristic index data of the line region within the preset time based on the power distribution network line loss characteristic index system, and specifically includes:

In a possible implementation manner, the power distribution network data processing module is configured to perform data processing on all power distribution network line loss data and all line loss characteristic index data, so as to obtain standard power distribution network line loss data and standard line loss characteristic index data, and specifically includes:

In a possible implementation manner, the power distribution network line loss prediction model building module is configured to build an LSTM neural network model based on an LSTM algorithm, perform model training on the LSTM neural network model according to the training data set and the test data set, and determine an optimal power distribution network line loss prediction model, which specifically includes:

In a possible implementation mode, the line loss characteristic index data in the power distribution network line loss characteristic index system building module comprises weather factors, equipment grid-connected parameters and distribution network operation data, wherein the weather factors comprise average temperature, average wind speed, illumination duration, illumination intensity and rainfall, the equipment grid-connected parameters comprise grid-connected capacity, grid-connected position, grid-connected operation mode and grid-connected electric quantity, and the distribution network operation data comprise data date, power type, power supply radius, line load rate and power supply quantity.

In a possible implementation manner, the line loss reasonable confidence interval calculation module is configured to acquire line loss characteristic index data to be predicted of different distributed power types, and input the line loss characteristic index data to be predicted to the optimal power distribution network line loss prediction model, and specifically includes:

In a possible implementation manner, the power distribution network line loss abnormal degree evaluation module is configured to calculate an abnormal coefficient of the suspected abnormal feeder line, and evaluate and obtain the power distribution network line loss abnormal degree of the suspected abnormal feeder line according to the abnormal coefficient, and specifically includes:

calculating the difference value between the suspected abnormal feeder line loss prediction data corresponding to the first preset time and the reference feeder line loss prediction data, and calculating the average difference value of all the difference values;

The invention also provides a terminal device, which comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to realize the method for evaluating the line loss abnormality of the feeder line of the power distribution network.

The invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program runs, a device in which the computer-readable storage medium is located is controlled to execute the method for evaluating the line loss abnormality of the feeder line of the power distribution network according to any one of the above items.

Compared with the prior art, the line loss abnormity evaluation method and device of the feeder line of the power distribution network have the following beneficial effects:

the method comprises the steps that a distribution network line loss characteristic index system containing the distributed power supply is constructed by considering line loss characteristic index data such as weather factors, equipment grid connection parameters and distribution network operation data, the influence analysis of the distributed power supply on line loss characteristics is achieved, the collected distribution network line loss data and line loss characteristic index data of a line area are used for training an LSTM neural network model, an optimal distribution network line loss prediction model is obtained, line loss prediction of the line loss characteristic index data to be predicted based on the optimal distribution network line loss prediction model is achieved, and a line loss reasonable confidence interval of the distribution network line loss is generated based on the first distribution network line loss prediction data; clustering the first power distribution network line loss prediction data based on the line loss reasonable confidence interval and an improved fuzzy C-means algorithm to obtain a suspected abnormal feeder, and evaluating the abnormal degree of the power distribution network line loss by calculating an abnormal coefficient of the suspected abnormal feeder. Compared with the prior art, the technical scheme of the invention breaks through the traditional method taking the line loss rate exceeding a certain threshold as the line loss abnormity evaluation condition, utilizes the long-time and short-time memory network LSTM and the improved fuzzy C mean value clustering algorithm to establish the abnormity coefficient to evaluate the line loss abnormity degree, can improve the accuracy of the line loss abnormity evaluation of the power distribution network, and simultaneously solves the problem that the dynamic change of the distributed power supply is not considered in the existing power distribution network line loss abnormity diagnosis.

Drawings

Fig. 1 is a schematic flow chart of an embodiment of a method for evaluating line loss abnormality of a feeder line of a power distribution network according to the present invention;

fig. 2 is a schematic structural diagram of an embodiment of a line loss abnormality evaluation device for a feeder line of a power distribution network provided by the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Example 1

Referring to fig. 1, fig. 1 is a schematic flowchart of an embodiment of a line loss abnormality evaluation method for a distribution network feeder line according to the present invention, as shown in fig. 1, the method includes steps 101 to 105, specifically as follows:

step 101: the method comprises the steps of obtaining line loss characteristic index data and constructing a power distribution network line loss characteristic index system containing the distributed power supply based on the line loss characteristic index data, wherein the line loss characteristic index data comprise weather factors, equipment grid-connected parameters and distribution network operation data.

In one embodiment, the distributed power supply types include wind power generation, photovoltaic power generation and hydroelectric power generation, and the obtained line loss characteristic index data are line loss characteristic index data corresponding to different distributed power supply types.

In one embodiment, the line loss characteristic index data includes weather factors, equipment grid-connected parameters and distribution network operation data, wherein the weather factors include average temperature, average wind speed, illumination duration, illumination intensity and rainfall, the equipment grid-connected parameters include grid-connected capacity, grid-connected position, grid-connected operation mode and grid-connected electric quantity, and the distribution network operation data includes data date, power type, power supply radius, line load rate and power supply quantity.

Specifically, for the selection of the weather factor, the main influence factor of wind power generation is wind energy density, and the wind energy density depends on air pressure, temperature and wind speed; the photovoltaic power generation is mainly determined by the total radiation amount of the sun, the total area of the photovoltaic cell and the photoelectric conversion efficiency, and under the condition that the conversion efficiency of the solar cell module is fixed, the power generation amount of a photovoltaic system is determined by the radiation intensity of the sun; the hydroelectric power generation mainly utilizes the water level changes of upstream and downstream, and the continuous high temperature or rainfall can cause certain influence on the water level changes, so that the hydroelectric power generation is influenced; therefore, through correlation analysis, the average temperature, the average wind speed, the illumination duration, the illumination intensity and the rainfall are determined as weather factors in the characteristic indexes.

Specifically, for the selection of equipment grid-connected parameters, the network structure, the operation mode, the trend direction and the like of the power distribution network can be changed due to the fact that the distributed power supply is connected in a grid mode, and main indexes of the equipment grid-connected parameter correlation of the distributed power supply are determined to be grid-connected capacity, grid-connected position, grid-connected operation mode and grid-connected electric quantity through analysis.

Specifically, for the selection of the distribution network operation data, data date, power type, power supply radius, line load rate and power supply amount are determined as the distribution network operation data based on big data analysis of main influence factors on the line loss of the distribution network.

In the embodiment, the output of the distributed power supply has uncertainty and randomness, so that the line loss change of the power distribution network can be influenced by the access of the distributed power supply, the line loss change trend is closely related to the characteristics of the distributed power supply, and the line loss characteristic index of the power distribution network containing the distributed power supply only considers a few factors and is not comprehensive at present, so that a power distribution network line loss characteristic index system containing the distributed power supply is established from 3 dimensions of weather factors, equipment grid-connected parameters and distribution network operation data, and the problem that the characteristic index of the power distribution network containing the distributed power supply is incomplete can be solved.

Step 102: collect all distribution network line loss data of the circuit region of choosing in the time of predetermineeing, and based on distribution network line loss characteristic index system collects all line loss characteristic index data of circuit region in the time of predetermineeing, it is right respectively all distribution network line loss data with all line loss characteristic index data carry out data processing, obtain standard distribution network line loss data and standard line loss characteristic index data, and will standard distribution network line loss data with standard line loss characteristic index data divide into training data set and test data set.

In one embodiment, the distributed power supply permeability of each line region in a power distribution network is obtained, the line region with the distributed power supply permeability larger than a preset distributed power supply permeability threshold is selected, and power distribution network line loss data corresponding to each day in one year of the line region are collected; wherein the distributed power supply permeability is a ratio of distributed power generation of the selected line to power supply of the selected line. Preferably, the preset distributed power supply permeability threshold is 50%.

In an embodiment, line loss characteristic index data corresponding to the line region in each day in the year are collected based on the power distribution network line loss characteristic index system, wherein weather factors in the line loss characteristic index data can be obtained from a Chinese meteorological bureau website, and equipment grid-connected parameters and power distribution network operation parameters in the line loss characteristic index data are obtained from a power distribution network management platform.

In an embodiment, after the line loss data of all the power distribution networks and the characteristic index data of all the line losses are obtained, data processing is performed on the line loss data of all the power distribution networks and the characteristic index data of all the line losses.

Specifically, negative values of the power distribution network line loss data in all the power distribution network line loss data are eliminated, and meanwhile, the power distribution network line loss data with the line loss rate larger than 25% in all the power distribution network line loss data are eliminated, so that first power distribution network line loss data are obtained.

Specifically, whether a data missing value exists in all the first power distribution network line loss data or all the line loss characteristic index data or not is further judged, if the data missing value exists, whether all the first power distribution network line loss data or all the line loss characteristic index data accord with uniform distribution or not is judged, if the data missing value exists, the mean value of all the first power distribution network line loss data or all the line loss characteristic index data is obtained, the mean value is filled in the data missing value, if the data missing value does not exist, the median of all the first power distribution network line loss data or all the line loss characteristic index data is obtained, and the median is filled in the data missing value, so that the standard power distribution network line loss data and the first line loss characteristic index data are obtained.

Specifically, because dimensions of characteristic indexes in the line loss characteristic index data are different, normalization processing needs to be performed on the first line loss characteristic index data, and the data is mapped to [0,1] to obtain standard loss characteristic index data, where a normalization processing formula is as follows:

in the formula, x _max Is the maximum value, x, of the single-class characteristic index data _min The single-type characteristic index data is the minimum value of the single-type characteristic index data, x is the original single-type characteristic index data to be normalized, and x' is the single-type characteristic index data after normalization, namely the single standard loss characteristic index data.

In one embodiment, the standard power distribution network line loss data and the standard line loss characteristic index data are divided into a training data set and a test data set, specifically, the standard power distribution network line loss data and the standard line loss characteristic index data are divided into the training data set and the test data set according to a division ratio by setting the division ratio; preferably, the division ratio is 8:2; and a group of standard distribution network line loss data and standard line loss characteristic index data in the training data set and the test data set are data acquired on the same day.

Step 103: and constructing an LSTM neural network model based on an LSTM algorithm, performing model training on the LSTM neural network model according to the training data set and the test data set, and determining an optimal power distribution network line loss prediction model.

In one embodiment, an LSTM neural network model is constructed based on an LSTM algorithm; specifically, an LSTM neural network model is constructed based on an LSTM algorithm, wherein the LSTM neural network model is a long-time memory neural network model and comprises 1 input layer, 1 hidden layer and 1 output layer, and the number of LSTM neurons in the hidden layer with the smallest error is searched by taking ReLu as an activation function; and an Adam algorithm is selected to solve the optimization problem of the line loss data and parameters of the power distribution network.

In one embodiment, weather factor influence indexes are determined according to distributed power types, namely three types of wind power generation, photovoltaic power generation and hydroelectric power generation, weather factors in characteristic index systems input by different distributed power types are different, when the distributed power type is a wind power type, average temperature and average wind speed are input, when the distributed power type is a photovoltaic type, illumination duration and illumination intensity are input, when the distributed power type is a hydraulic type, average temperature and rainfall are input, and equipment grid-connected parameters and power distribution network operation data characteristic indexes are input by the three distributed power types and serve as input of a neural network model.

In one embodiment, the training data set is input into the LSTM neural network model for model training to obtain an initial power distribution network line loss prediction model, wherein the initial power distribution network line loss prediction model is an initial power distribution network line loss prediction model containing a distributed power supply; and inputting the test data set into the initial power distribution network line loss prediction model to obtain power distribution network line loss prediction data corresponding to the test data set.

In one embodiment, the power distribution network line loss prediction data is compared with the standard power distribution network line loss data to determine an optimal power distribution network line loss prediction model; specifically, the power distribution network line loss prediction data corresponding to the test data set is compared with the standard power distribution network line loss data in the test set, that is, the predicted value is compared with the actual value for analysis, and the prediction result is evaluated; calculating the mean square error and the mean absolute error of the power distribution network line loss prediction data corresponding to the test data set and the standard power distribution network line loss data in the test set, and determining an optimal power distribution network line loss prediction model when the mean square error and the mean absolute error are minimum, wherein the calculation formulas of the mean absolute error (MAPE) and the Mean Square Error (MSE) are as follows:

′

in the formula, y _i The method comprises the steps that the ith power distribution network line loss prediction data in a test data set is obtained; y is _i The method comprises the steps that the line loss data of the ith standard distribution network in a test set are tested; and n is the number of samples in the test set, namely the number of power distribution network lines.

Step 104: obtaining characteristic index data of the line loss to be predicted of different distributed power types, inputting the characteristic index data of the line loss to be predicted into the optimal power distribution network line loss prediction model, outputting and calculating line loss reasonable confidence intervals of the power distribution network line losses of different distributed power types according to first power distribution network line loss prediction data corresponding to different distributed power types.

In one embodiment, line loss characteristic index data to be predicted of different distributed power supply types is obtained; specifically, first to-be-predicted line loss characteristic index data of a distributed power supply type for wind power generation is obtained, wherein weather factors in the first to-be-predicted line loss characteristic index data comprise average temperature and average wind speed; acquiring second line loss characteristic index data to be predicted, wherein the type of the distributed power supply is photovoltaic power generation, and weather factors in the second line loss characteristic index data to be predicted comprise illumination duration and illumination intensity; and acquiring third line loss characteristic index data to be predicted, wherein the distributed power supply type is hydroelectric power generation, and weather factors in the third line loss characteristic index data to be predicted comprise average temperature and rainfall.

In an embodiment, the first to-be-predicted line loss characteristic index data, the second to-be-predicted line loss characteristic index data and the third to-be-predicted line loss characteristic index data are input into the optimal power distribution network line loss prediction model, so that the optimal power distribution network line loss prediction model outputs first power distribution network line loss prediction data corresponding to different distributed power supply types.

In one embodiment, line loss reasonable confidence intervals of the line losses of the power distribution networks of different distributed power supply types are calculated according to the line loss prediction data of the first power distribution network corresponding to the different distributed power supply types.

Specifically, according to the first distribution network line loss prediction data corresponding to different distributed power supply types, confidence intervals of the first distribution network line loss prediction data corresponding to different distributed power supply types when the confidence degrees are 90%, 95% and 99% are respectively calculated, wherein the confidence intervals are intervals formed by taking a confidence upper limit and a confidence lower limit of statistics as an upper bound and a lower bound respectively, and the confidence intervals are as follows:

in the formula (I), the compound is shown in the specification,

the average value of the line loss prediction data of the first power distribution network is obtained; the C value at 90% confidence is 1.64, the C value at 95% confidence is 1.96, and the C value at 99% confidence is 2.58; σ is the standard deviation of the overall data; and n is the number of the first distribution network line loss prediction data.

In one embodiment, the confidence interval under 95% confidence is selected as the line loss reasonable confidence interval of the line loss of the power distribution network.

Step 105: the method comprises the steps of carrying out primary clustering on first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, carrying out secondary clustering on the primary clustering result based on a line loss reasonable confidence interval to obtain a suspected abnormal feeder, calculating an abnormal coefficient of the suspected abnormal feeder, and evaluating and obtaining the power distribution network line loss abnormal degree of the suspected abnormal feeder according to the abnormal coefficient.

In an embodiment, a fuzzy coefficient and an iteration stop threshold in the improved fuzzy C-means algorithm are set, and the first power distribution network line loss prediction data obtained in step 104 and the equipment grid-connected parameters and power distribution network operation data in the line loss characteristic index data to be predicted corresponding to the first power distribution network line loss prediction data are used as input of the improved fuzzy C-means clustering algorithm, so that the improved fuzzy C-means clustering algorithm clusters the first power distribution network line loss prediction data. Preferably, the blurring coefficient m =2 is set, and the threshold for stopping the iteration is set to 0.001.

In one embodiment, the clustering process is a process of minimizing a target function, an error value of the target function is gradually reduced through repeated iterative operation, and when the target function is converged, a final primary clustering result can be obtained; wherein, the clustering process is as follows:

order to

The given number of samples is N, the spatial dimension of the samples is S, and the number of clusters is set as C. The objective function is:

/>

such that:

u _ij ≥0，1≤i≤N，1≤j≤C；

in the formula, U is a membership matrix; v is a matrix consisting of C cluster centers; u. of _ij The membership degree of the ith sample to the jth class; the degree of membership represents the degree to which an object x is subject to the set Q, with closer to 1 representing higher degrees of membership. x is the number of _i Is the ith sample; v. of _j Is the jth cluster center; m is a blurring coefficient; | x _i -v _j | | denotes the sample point x _i To the center of the cluster v _j The euclidean distance of (c).

In one embodiment, the optimal clustering number is selected through the effectiveness function, and the Xie-Beni clustering effectiveness function is selected to determine the optimal classification number.

The XB index can find a certain balance point between the intra-class compactness and the inter-class separation, and the formula is as follows:

in the formula: x is the number of _i Is the ith sample; v. of _j Is the jth cluster center; v. of _i Is the ith cluster center.

Wherein, in-class compactness = formula molecular fraction/sample number N, the smaller the better; the degree of separation between classes = the number of samples N of the denominator part x of the formula, the larger the better; the smaller the calculation result of the XB index is, the better the clustering effect is.

In one embodiment, the primary clustering result is subjected to secondary clustering based on the line loss reasonable confidence interval to obtain a suspected abnormal feeder line.

Specifically, the clustering centers in the primary clustering and the classes thereof are analyzed based on the end points of the line loss reasonable confidence intervals; and acquiring the class of the clustering center outside the line loss reasonable confidence interval and less than 0.5 of the endpoint of the line loss reasonable confidence interval, performing secondary clustering on the class of the clustering center, screening out data in the line loss reasonable confidence interval of the clustering center through the secondary clustering, and determining the class of the clustering center outside the line loss reasonable confidence interval as a suspected abnormal feeder.

In one embodiment, in order to quantitatively analyze the abnormal degree of the feeder line loss, from line loss space-time distribution feature analysis, the abnormal coefficient of the suspected abnormal feeder line is calculated by taking the clustering center of the category where the suspected abnormal feeder line is located as a reference feeder line and considering the time dispersion aspect and the space dispersion aspect.

Specifically, with a clustering center of a category of suspected abnormal feeder line loss as a reference feeder line, acquiring suspected abnormal feeder line loss prediction data and suspected abnormal feeder line loss characteristic index data of the suspected abnormal feeder line, and acquiring reference feeder line loss prediction data and reference feeder line loss characteristic index data of the reference feeder line at the same time;

specifically, a difference value between the suspected abnormal feeder line loss prediction data corresponding to the first preset time and the reference feeder line loss prediction data is calculated, and a difference value average value of all the difference values is calculated; calculating the deviation between the suspected abnormal feeder line loss characteristic index data and the reference feeder line loss characteristic index data, and calculating the average deviation value corresponding to each line loss characteristic index based on the number of the line loss characteristic indexes; the calculation process of the difference average value A and the deviation average value B is as follows:

in the formula, T is the number of days of a first preset time; x _i Predicting data for the line loss of the ith suspected abnormal feeder line; x _c Is the reference feeder line loss; s is the number of line loss characteristic indexes; z _i Characteristic index data of the line loss of the ith suspected abnormal feeder line are obtained; z is a linear or branched member _c The characteristic index data of the reference feeder line.

Preferably, the first preset time is the same time interval of each day in one month, that is, a difference between the suspected abnormal feeder line loss prediction data and the reference feeder line loss prediction data corresponding to the same time interval of each day in a preset month is calculated.

Specifically, the difference average value and the deviation average value are added according to the difference average value and the deviation average value to obtain an abnormal coefficient F of the suspected abnormal feeder line, where F = a + B.

In one embodiment, the abnormal coefficient is compared with a preset abnormal coefficient threshold, and if the abnormal coefficient is greater than or equal to the preset abnormal coefficient threshold, it is determined that the distribution network line loss of the suspected abnormal feeder line is abnormal to a higher degree, that is, the feeder line has a higher possibility of having abnormal line loss in a certain time period; if the abnormal coefficient is smaller than the preset abnormal coefficient threshold, the distribution network line loss abnormal degree of the suspected abnormal feeder line is considered to be low, namely the possibility that the feeder line has abnormal line loss in a certain time period is low.

Example 2

Referring to fig. 2, fig. 2 is a schematic structural diagram of an embodiment of an apparatus for evaluating line loss abnormality of a feeder line of a power distribution network, as shown in fig. 2, the apparatus includes a power distribution network line loss characteristic index system building module 201, a power distribution network data processing module 202, a power distribution network line loss prediction model building module 203, a line loss reasonable confidence interval calculating module 204, and a power distribution network line loss abnormality degree evaluating module 205, which are specifically as follows:

the power distribution network line loss characteristic index system construction module 201 is used for acquiring and constructing a power distribution network line loss characteristic index system containing a distributed power supply based on line loss characteristic index data, wherein the line loss characteristic index data comprise weather factors, equipment grid-connected parameters and distribution network operation data.

Distribution network data processing module 202 for collect all distribution network line loss data of the circuit region of selecting in the time of predetermineeing, and be based on distribution network line loss characteristic index system collects all line loss characteristic index data of circuit region in the time of predetermineeing, it is right respectively all distribution network line loss data with all line loss characteristic index data carry out data processing, obtain standard distribution network line loss data and standard line loss characteristic index data, and will standard distribution network line loss data with standard line loss characteristic index data divide into training data set and test data set.

The power distribution network line loss prediction model construction module 203 is configured to construct an LSTM neural network model based on an LSTM algorithm, perform model training on the LSTM neural network model according to the training data set and the test data set, and determine an optimal power distribution network line loss prediction model.

The line loss reasonable confidence interval calculation module 204 is configured to obtain line loss characteristic index data to be predicted of different distributed power types, input the line loss characteristic index data to be predicted to the optimal power distribution network line loss prediction model, output and calculate line loss reasonable confidence intervals of power distribution network line losses of different distributed power types according to first power distribution network line loss prediction data corresponding to different distributed power types.

The power distribution network line loss abnormal degree evaluation module 205 is configured to perform primary clustering on first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, perform secondary clustering on the primary clustering result based on the line loss reasonable confidence interval to obtain a suspected abnormal feeder line, calculate an abnormal coefficient of the suspected abnormal feeder line, and evaluate and obtain the power distribution network line loss abnormal degree of the suspected abnormal feeder line according to the abnormal coefficient.

In an embodiment, the power distribution network data processing module 202 is configured to collect all power distribution network line loss data of the selected line region in a preset time, and collect all line loss characteristic index data of the line region in the preset time based on the power distribution network line loss characteristic index system, and specifically includes: the method comprises the steps of obtaining the permeability of the distributed power supply of each line area in the power distribution network, selecting the line area of which the permeability is larger than a preset permeability threshold of the distributed power supply, and collecting line loss data of the power distribution network of the line area within one year; and collecting line loss characteristic index data of the line region based on the power distribution network line loss characteristic index system.

In an embodiment, the power distribution network data processing module 202 is configured to perform data processing on the all power distribution network line loss data and the all line loss characteristic index data, respectively, to obtain standard power distribution network line loss data and standard line loss characteristic index data, and specifically includes: removing negative values of the power distribution network line loss data in all the power distribution network line loss data, and removing the power distribution network line loss data with the line loss rate of more than 25% in all the power distribution network line loss data to obtain first power distribution network line loss data; judging whether data missing values exist in all the first power distribution network line loss data or all the line loss characteristic index data or not, if so, judging whether all the first power distribution network line loss data or all the line loss characteristic index data are in accordance with uniform distribution or not, if so, acquiring the mean value of all the first power distribution network line loss data or all the line loss characteristic index data and filling the mean value into the data missing values, otherwise, acquiring the median of all the first power distribution network line loss data or all the line loss characteristic index data and filling the median into the data missing values to obtain standard power distribution network line loss data and first line loss characteristic index data; and carrying out normalization processing on the first line loss characteristic index data to obtain standard line loss characteristic index data.

In an embodiment, the power distribution network line loss prediction model building module 203 is configured to build an LSTM neural network model based on an LSTM algorithm, perform model training on the LSTM neural network model according to the training data set and the test data set, and determine an optimal power distribution network line loss prediction model, which specifically includes: constructing an LSTM neural network model based on an LSTM algorithm, inputting the training data set into the LSTM neural network model for model training to obtain an initial power distribution network line loss prediction model, and inputting the test data set into the initial power distribution network line loss prediction model to obtain power distribution network line loss prediction data corresponding to the test data set; and comparing the power distribution network line loss prediction data with the standard power distribution network line loss data to determine an optimal power distribution network line loss prediction model.

In an embodiment, in the power distribution network line loss characteristic index system building module 201, the line loss characteristic index data includes a weather factor, an equipment grid-connected parameter and distribution network operation data, wherein the weather factor includes an average temperature, an average wind speed, illumination duration, illumination intensity and rainfall, the equipment grid-connected parameter includes a grid-connected capacity, a grid-connected position, a grid-connected operation mode and an internet electric quantity, and the distribution network operation data includes a data date, a power type, a power supply radius, a line load rate and a power supply quantity.

In an embodiment, the line loss reasonable confidence interval calculation module 204 is configured to obtain line loss characteristic index data to be predicted of different distributed power source types, and input the line loss characteristic index data to be predicted to the optimal power distribution network line loss prediction model, and specifically includes: acquiring first to-be-predicted line loss characteristic index data of a distributed power supply type for wind power generation, wherein weather factors in the first to-be-predicted line loss characteristic index data comprise average temperature and average wind speed; acquiring second line loss characteristic index data to be predicted, wherein the type of the distributed power supply is photovoltaic power generation, and weather factors in the second line loss characteristic index data to be predicted comprise illumination duration and illumination intensity; acquiring third line loss characteristic index data to be predicted, wherein the distributed power supply is in a hydroelectric power generation type, and weather factors in the third line loss characteristic index data to be predicted comprise average temperature and rainfall; and inputting the first to-be-predicted line loss characteristic index data, the second to-be-predicted line loss characteristic index data and the third to-be-predicted line loss characteristic index data into the optimal power distribution network line loss prediction model.

In an embodiment, the power distribution network line loss abnormal degree evaluation module 205 is configured to calculate an abnormal coefficient of the suspected abnormal feeder, and evaluate and obtain the power distribution network line loss abnormal degree of the suspected abnormal feeder according to the abnormal coefficient, and specifically includes: taking a clustering center of the category of the suspected abnormal feeder line loss as a reference feeder line, acquiring suspected abnormal feeder line loss prediction data and suspected abnormal feeder line loss characteristic index data of the suspected abnormal feeder line, and acquiring reference feeder line loss prediction data and reference feeder line loss characteristic index data of the reference feeder line; calculating the difference value between the suspected abnormal feeder line loss prediction data corresponding to the first preset time and the reference feeder line loss prediction data, and calculating the average difference value of all the difference values; calculating the deviation between the suspected abnormal feeder line loss characteristic index data and the reference feeder line loss characteristic index data, and calculating the deviation average value corresponding to each line loss characteristic index based on the number of the line loss characteristic indexes; calculating and obtaining an abnormal coefficient of the suspected abnormal feeder line according to the difference average value and the deviation average value; and comparing the abnormal coefficient with a preset abnormal coefficient threshold, if the abnormal coefficient is greater than or equal to the preset abnormal coefficient threshold, considering that the distribution network line loss abnormal degree of the suspected abnormal feeder line is higher, and if the abnormal coefficient is less than the preset abnormal coefficient threshold, considering that the distribution network line loss abnormal degree of the suspected abnormal feeder line is lower.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working process of the apparatus described above may refer to the corresponding process in the foregoing method embodiment, and is not described herein again.

It should be noted that the above-mentioned embodiment of the device for evaluating the line loss abnormality of the feeder line of the distribution network is merely illustrative, where the modules described as the separate components may or may not be physically separate, and the components displayed as the modules may or may not be physical units, that is, may be located in one place, or may also be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.

On the basis of the above embodiment of the line loss abnormality evaluation method for the power distribution network feeder line, another embodiment of the present invention provides a line loss abnormality evaluation terminal device for the power distribution network feeder line, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the line loss abnormality evaluation method for the power distribution network feeder line according to any one of the embodiments of the present invention is implemented.

Illustratively, the computer program may be partitioned in this embodiment into one or more modules that are stored in the memory and executed by the processor to implement the invention. The one or more modules can be a series of instruction segments of a computer program capable of performing specific functions, and the instruction segments are used for describing the execution process of the computer program in the line loss abnormity evaluation terminal equipment of the distribution network feeder line.

The line loss abnormity evaluation terminal equipment of the power distribution network feeder line can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing equipment. The line loss abnormity evaluation terminal device of the power distribution network feeder line can comprise, but is not limited to, a processor and a memory.

The processor may be a Central Processing Unit (CPU), other general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general processor may be a microprocessor or the processor may be any conventional processor, etc., and the processor is a control center of the line loss abnormality evaluation terminal device of the distribution network feeder line, and various interfaces and lines are used to connect various parts of the line loss abnormality evaluation terminal device of the whole distribution network feeder line.

The memory can be used for storing the computer programs and/or modules, and the processor realizes various functions of the line loss abnormality evaluation terminal device of the power distribution network feeder line by running or executing the computer programs and/or modules stored in the memory and calling the data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the mobile phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.

On the basis of the above embodiment of the line loss anomaly evaluation method for the feeder line of the power distribution network, another embodiment of the present invention provides a storage medium, where the storage medium includes a stored computer program, and when the computer program runs, a device on which the storage medium is located is controlled to execute the line loss anomaly evaluation method for the feeder line of the power distribution network according to any embodiment of the present invention.

In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in source code form, object code form, an executable file or some intermediate form, and so on. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, read-Only Memory (ROM), random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer-readable medium may contain suitable additions or subtractions depending on the requirements of legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer-readable media may not include electrical carrier signals or telecommunication signals in accordance with legislation and patent practice.

In summary, the invention discloses a line loss abnormity evaluation method and device of a power distribution network feeder line, and the invention discloses the line loss abnormity evaluation method and device of the power distribution network feeder line, wherein a power distribution network line loss characteristic index system containing a distributed power supply is constructed, and the collected power distribution network line loss data and line loss characteristic index data of a line area are used for training an LSTM neural network model to obtain an optimal power distribution network line loss prediction model, so that the line loss prediction of the line loss characteristic index data to be predicted based on the optimal power distribution network line loss prediction model is obtained and based on first power distribution network line loss prediction data, and a line loss reasonable confidence interval of the power distribution network line loss is generated; clustering the first power distribution network line loss prediction data based on the line loss reasonable confidence interval and an improved fuzzy C-means algorithm to obtain a suspected abnormal feeder, and evaluating the abnormal degree of the power distribution network line loss by calculating an abnormal coefficient of the suspected abnormal feeder. Compared with the prior art, the technical scheme of the invention breaks through the traditional method taking the line loss rate exceeding a certain threshold value as the line loss abnormity evaluation condition, establishes the abnormity coefficient to evaluate the line loss abnormity degree, and can improve the accuracy of the line loss abnormity evaluation of the power distribution network.

The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and substitutions can be made without departing from the technical principle of the present invention, and these modifications and substitutions should also be regarded as the protection scope of the present invention.

Claims

1. A line loss abnormity evaluation method of a power distribution network feeder line is characterized by comprising the following steps:

collecting all power distribution network line loss data of a selected line region within preset time, collecting all line loss characteristic index data of the line region within the preset time based on a power distribution network line loss characteristic index system, respectively carrying out data processing on all power distribution network line loss data and all line loss characteristic index data to obtain standard power distribution network line loss data and standard line loss characteristic index data, and dividing the standard power distribution network line loss data and the standard line loss characteristic index data into a training data set and a testing data set;

the method comprises the steps of carrying out primary clustering on first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, carrying out secondary clustering on the primary clustering result based on a line loss reasonable confidence interval to obtain a suspected abnormal feeder, calculating an abnormal coefficient of the suspected abnormal feeder, and evaluating and obtaining the power distribution network line loss abnormal degree of the suspected abnormal feeder according to the abnormal coefficient.

2. The method for evaluating the line loss abnormality of the distribution network feeder line according to claim 1, wherein the step of collecting all distribution network line loss data of the selected line region within a preset time, and collecting all line loss characteristic index data of the line region within the preset time based on the distribution network line loss characteristic index system specifically comprises:

3. The method according to claim 1, wherein the data processing is performed on the all-distribution-network line loss data and the all-line loss characteristic index data to obtain standard distribution-network line loss data and standardized line loss characteristic index data, and specifically includes:

removing negative values of the power distribution network line loss data in all the power distribution network line loss data, and removing the power distribution network line loss data with the line loss rate larger than 25% in all the power distribution network line loss data to obtain first power distribution network line loss data;

4. The method for evaluating the line loss abnormality of the power distribution network feeder line according to claim 1, wherein an LSTM neural network model is constructed based on an LSTM algorithm, model training is performed on the LSTM neural network model according to the training data set and the test data set, and an optimal power distribution network line loss prediction model is determined, specifically comprising:

5. The method according to claim 1, wherein the line loss characteristic index data comprises weather factors, equipment grid-connected parameters and distribution network operation data, wherein the weather factors comprise average temperature, average wind speed, illumination duration, illumination intensity and rainfall, the equipment grid-connected parameters comprise grid-connected capacity, grid-connected position, grid-connected operation mode and grid-connected electric quantity, and the distribution network operation data comprises data date, power type, power supply radius, line load rate and power supply quantity.

6. The method for evaluating the line loss abnormality of the power distribution network feeder line according to claim 5, wherein the steps of obtaining line loss characteristic index data to be predicted of different distributed power types, and inputting the line loss characteristic index data to be predicted into the optimal power distribution network line loss prediction model specifically comprise:

and inputting the first to-be-predicted line loss characteristic index data, the second to-be-predicted line loss characteristic index data and the third to-be-predicted line loss characteristic index data into the optimal power distribution network line loss prediction model.

7. The method according to claim 1, wherein the step of calculating an abnormal coefficient of the suspected abnormal feeder line, and estimating and obtaining the abnormal degree of the distribution network line loss of the suspected abnormal feeder line according to the abnormal coefficient includes:

8. A line loss abnormity evaluation device of a power distribution network feeder line is characterized by comprising: the system comprises a power distribution network line loss characteristic index system building module, a power distribution network data processing module, a power distribution network line loss prediction model building module, a line loss reasonable confidence interval calculating module and a power distribution network line loss abnormal degree evaluating module;

the power distribution network line loss abnormal degree evaluation module is used for carrying out primary clustering on the first power distribution network line loss prediction data based on an improved fuzzy C-means algorithm to obtain a primary clustering result, carrying out secondary clustering on the primary clustering result based on the line loss reasonable confidence interval to obtain a suspected abnormal feeder, calculating an abnormal coefficient of the suspected abnormal feeder, and evaluating and obtaining the power distribution network line loss abnormal degree of the suspected abnormal feeder according to the abnormal coefficient.

9. A terminal device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor when executing the computer program implementing the method for line loss anomaly assessment for a distribution feeder according to any one of claims 1 to 7.

10. A computer-readable storage medium, comprising a stored computer program, wherein when the computer program runs, the computer-readable storage medium controls a device to execute the method for evaluating line loss abnormality of a feeder line of a power distribution network according to any one of claims 1 to 7.