Model mechanism and data driving combined asynchronous fault location method for double-end long-distance power transmission line
By using model mechanism and data-driven methods in long-distance transmission lines, a double-ended equivalent model is built and a neural network is used for fault ranging, the ranging error problem caused by uneven distribution of line parameters in traditional methods is solved, and the realization of the realization of the realization of the distance measurement error problem. High-precision and robust fault ranging.
Patent Information
- Application Number
- CN202410859292.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2024-11-12
AI Technical Summary
The traditional uniform transmission line fault ranging algorithm in long-distance transmission lines has large distance measurement errors due to uneven distribution of line parameters. The existing methods rely on specific line models and parameters, which are easily affected by errors and are difficult to accurately measure ranging.
Using a dual-end long-distance transmission line asynchronous fault ranging method combined with model mechanism and data drive, fault ranging is realized by building a dual-end equivalent model and a data-driven neural network model, and avoiding the line model and parameters. rely.
This method can measure the distance with high accuracy without relying on line models and parameters, reducing the impact on fault type, position, transition resistance and other factors, and improving the accuracy and robustness of the distance measurement.
Smart Images

Figure CN118937884A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of relay protection of electric power systems, and in particular relates to a method for asynchronous fault location of a double-terminal long-distance transmission line combining a model mechanism with data drive. Background Art
[0002] Accurate fault distance measurement algorithms are conducive to the timely repair of transmission lines and the rapid recovery of power supply systems. They have always been an important research topic in power systems. A large number of fault distance measurement algorithms have been proposed for uniform transmission lines. With the development of my country's power system, the application of long-distance transmission lines has become more and more extensive due to the advantages of low transmission loss and high power, and has become a common transmission method. Long-distance transmission lines are affected by soil, climate and other factors in various regions, resulting in uneven distribution of line parameters. Traditional uniform transmission line fault distance measurement algorithms will produce large distance measurement errors in long-distance transmission lines. Therefore, it is also necessary to study fault distance measurement methods for long-distance transmission lines.
[0003] The existing partial hybrid line fault distance measurement method has been extended to the n-segment cable-overhead line hybrid line. Although this method can be used for reference in long-distance transmission lines with uneven parameter distribution, there is a problem that the line segmentation and parameters of each section of the long-distance transmission line are difficult to determine. Existing fault distance measurement methods are mostly based on specific line models and accurate line parameters, and the distance measurement accuracy is easily affected by line model and parameter errors. Therefore, considering the problem of uneven parameter distribution of long-distance transmission lines, it is of practical significance to propose a transmission line fault distance measurement method that does not require line models and line parameters. With the widespread application of artificial intelligence technology in power systems, artificial intelligence algorithms also provide a solution to the fault distance measurement problem of power lines without the need for line models and line parameters. Summary of the invention
[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a two-terminal long-distance transmission line asynchronous fault location method that combines a model mechanism with data drive. The method takes into account the problem that long-distance transmission lines are affected by factors such as soil and climate in various regions, resulting in uneven distribution of line parameters, and effectively makes up for the shortcomings of the existing methods.
[0005] The present invention solves the technical problem by the following technical solutions:
[0006] A method for non-synchronous fault location of a double-terminal long-distance transmission line combining a model mechanism with data drive, the method comprising the following steps:
[0007] Construct an equivalent model of a two-terminal long-distance transmission line;
[0008] The equivalent model mechanism analysis of double-terminal long-distance transmission lines is carried out under the distributed parameter model and Π-type line model;
[0009] Acquire data and use data-driven fault location measurement for double-terminal long-distance transmission lines.
[0010] Moreover, the equivalent model of the two-terminal long-distance transmission line takes into account that the two-terminal long-distance transmission line is affected by soil and climate factors in various regions, resulting in uneven distribution of line parameters. The two-terminal long-distance transmission line is regarded as a non-uniform transmission line connected by n sections of lines with different parameters and lengths. The length of the i-th section is l i , the total length of the transmission line is L n .
[0011] Moreover, each line segment of the distributed parameter model is equivalent to the distributed parameter model, and the positive sequence wave impedance and positive sequence propagation constant of the i-th line segment are Z ci and γ i ;
[0012] Each section of the Π-type line model is equivalent to a Π-type line model, and the positive sequence impedance and positive sequence admittance of the i-th section of the line are Z i and Y i .
[0013] Moreover, the analysis of the mechanism of the equivalent model of the two-terminal long-distance transmission line under the distributed parameter model is specifically as follows:
[0014] When a two-terminal long-distance transmission line operates normally, the relationship between the positive sequence voltage and positive sequence current at both ends of the line is as follows:
[0015]
[0016] in: and They are the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal before the line fault respectively;
[0017] δ is the asynchronous angle of the data collected by the M end relative to the data collected by the N end, taking into account the asynchronous sampling of the data at both ends;
[0018] If a short circuit fault occurs in a double-terminal long-distance transmission line, the distance measurement function of the double-terminal long-distance transmission line based on the distributed parameter model is derived according to the fault analysis:
[0019]
[0020] in: and They are the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal after the line fault respectively;
[0021] Moreover, the equivalent model mechanism analysis of the double-terminal long-distance transmission line under the Π-type line model is specifically as follows:
[0022] When a two-terminal long-distance transmission line operates normally, the relationship between the positive sequence voltage and positive sequence current at both ends of the line is:
[0023]
[0024] If a double-terminal long-distance transmission line fails, after fault analysis and formula derivation, the distance measurement function of the double-terminal long-distance transmission line based on the Π-type line model can be obtained as follows:
[0025]
[0026]
[0027] Furthermore, the acquisition of data and the use of data-driven fault location measurement for a two-terminal long-distance power transmission line are specifically as follows:
[0028] Acquire data, obtain the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal before and after the line fault;
[0029] The method of using data-driven fault location measurement for a two-terminal long-distance transmission line includes: preprocessing data, designing and training a two-terminal long-distance transmission line fault location measurement model, and performing fault location measurement according to the trained model;
[0030] The data is preprocessed to collect the positive sequence voltage and current at the M end and the positive sequence voltage and current at the N end before the two sets of line faults, and the A in formula (1) is obtained by matrix calculation. n A n-1 …A1e jδ Or B in formula (6) n B n-1 …B1e jδ ;
[0031] Substitute the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal after the line fault into equation (4) or equation (9), and we can get The real part k1 and imaginary part k2 of are used as the line electrical characteristic quantities after data preprocessing;
[0032] Designing and training a two-terminal long-distance transmission line fault location model includes designing a three-layer neural network model, using k1 and k2 as inputs of the neural network, and the corresponding fault location as output of the neural network. The present invention comprehensively considers output error and overfitting phenomenon, selects the number of hidden layer neurons as 10, and the activation function is the tansig function;
[0033] The network is trained using historical fault data to fit the distance measurement function in equation (3) or equation (8). After the training is completed, the functional relationship between the line electrical characteristic quantity and the fault location is stored in the neural network in the form of connection weights and thresholds between neurons.
[0034] Fault distance measurement based on the trained model includes collecting line fault data for preprocessing when a line fault occurs, inputting the obtained line electrical characteristic quantity into the trained fault distance measurement model, and the output of the model is the corresponding fault location.
[0035] The advantages and beneficial effects of the present invention are:
[0036] The method of the present invention establishes a fault distance measurement model of a double-terminal long-distance transmission line by analyzing the model mechanism of the double-terminal long-distance transmission line and utilizing data-driven. The distance measurement process does not rely on the line model and does not require line parameters. In the modeling process, feature transformation and feature aggregation of fault data are realized through data preprocessing, and the preprocessed data is input into the neural network for training, so that the neural network training does not require a large number of historical fault samples. The distance measurement accuracy of the method of the present invention is not affected by factors such as fault type, fault location, transition resistance, fault initial phase angle, load current, and asynchronous data. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 A schematic diagram of a system when a fault occurs in a double-terminal long-distance power transmission line applied to the present invention;
[0038] Figure 2 A schematic diagram of a neural network structure for double-terminal long-distance power transmission line fault location in the present invention;
[0039] Figure 3 The figure is a schematic diagram of the overall process of the double-terminal long-distance power transmission line fault location method in the present invention. DETAILED DESCRIPTION
[0040] The present invention is further described in detail below through specific examples. The following examples are only illustrative and not restrictive, and the protection scope of the present invention cannot be limited thereto.
[0041] Figure 1 The figure is a fault diagram of a two-terminal long-distance transmission line applied to the present invention. The two-terminal long-distance transmission line can be regarded as a non-uniform transmission line that connects n lines with different parameters and lengths. In the figure, the distance between the fault point f and the M end is x, and the length of the i-th (i=1, 2, ..., n) line is l i , the total length of the transmission line is L nThe present invention provides a method for asynchronous fault location of a double-terminal long-distance power transmission line combining a model mechanism with data drive, which mainly includes a model mechanism analysis of a double-terminal long-distance power transmission line and a fault location method of a double-terminal long-distance power transmission line.
[0042] 1. Model mechanism analysis of double-terminal long-distance transmission lines
[0043] All the analyses in this invention take into account the asynchrony of the data sampling at both ends. Taking the time of data collection at the N end as a reference, the asynchrony angle of data collection at the M end relative to data collection at the N end is δ. That is, when considering the asynchrony of data collection at both ends, if the positive sequence voltage and positive sequence current at the N end are respectively and Then the positive sequence voltage and positive sequence current at terminal M are and
[0044] (1) Analysis of fault location of two-terminal long-distance transmission lines based on distributed parameter model
[0045] The positive-sequence wave impedance and positive-sequence propagation constant of the i-th (i=1, 2, ..., n) line are Z ci and γ i When a fault occurs at point f in the kth line, according to the uniform transmission line equation, the positive sequence voltage at the M end is and positive sequence current Calculate the positive sequence voltage at the fault point and the positive sequence current flowing into the fault point from the M side for:
[0046]
[0047] Where, L k-1 is the total length of the first k-1 segments of the line. Similarly, L in the following text k , L k+1 are the total lengths of the first k and k+1 segments of the line respectively.
[0048] The positive sequence voltage at the fault point and the positive sequence current flowing from the fault point to the N side Calculate the positive sequence voltage at the N terminal and positive sequence current for:
[0049]
[0050] according to in is the positive sequence current flowing into the fault point, and it can be obtained from formula (12):
[0051]
[0052] Substituting formula (11) into formula (13), we can obtain:
[0053]
[0054] Where:
[0055]
[0056] make
[0057]
[0058] Simplifying formula (14) we can get:
[0059]
[0060] Where:
[0061]
[0062] Divide both sides of the equal sign in equation (17) to eliminate the unknown quantity. Then, after equation transformation, the distance measurement equation can be obtained as follows:
[0063]
[0064] When the double-terminal long-distance transmission line operates normally, the positive sequence voltage at the M terminal is Positive sequence current and the positive sequence voltage at the N terminal Positive sequence current The relationship is as follows:
[0065]
[0066] (2) Fault location analysis of two-terminal long-distance transmission lines based on the Π-type line model
[0067] Each section of the double-terminal long-distance transmission line is equivalent to a Π-type model. The positive-sequence impedance and positive-sequence admittance of the i-th (i=1, 2, …, n) section of the line are Z i and Y i , similar to the above analysis, it can be deduced that when the double-terminal long-distance transmission line operates normally, the relationship between the positive sequence voltage and positive sequence current at both ends of the line is:
[0068]
[0069] Where:
[0070]
[0071] When a double-terminal long-distance transmission line fails at point f in the kth section, let
[0072]
[0073] After a series of derivations, the distance measurement function of the two-terminal long-distance transmission line based on the Π-type line model can be obtained as follows:
[0074]
[0075] Where:
[0076]
[0077] 2. Fault location method for double-terminal long-distance transmission lines
[0078] According to equations (16), (20) or (21), (23), the positive sequence voltage and current at both ends of the line before and after the fault are used to obtain the distance function From equations (19) and (20), we can see that the distance measurement function of a two-terminal long-distance transmission line is a piecewise function, which requires the number of line segments n and the parameters Z of each line segment to be given in advance. ci , γ i , l i or Z i , Y i , l i In fact, the number of line sections and the parameters of each line section of a two-terminal long-distance transmission line may be unknown or inaccurate, so it is difficult to directly construct and solve the ranging function. Moreover, f(x) constructed based on different line models is different. Directly using the above ranging function to calculate the fault location is easily affected by line modeling, resulting in inaccurate ranging results.
[0079] After analyzing and summarizing the distance measurement principle of the double-terminal long-distance transmission line, the method of the present invention pre-processes the original voltage and current at both ends of the line, and then uses the historical fault data of the transmission line and the function fitting ability of the neural network to establish a fault distance measurement model for the double-terminal long-distance transmission line. The structure of the neural network used in the present invention is as follows: Figure 2 As shown, the overall process of the double-terminal transmission line fault location algorithm of the present invention is as follows Figure 3 As shown, the specific steps are as follows:
[0080] (1) The positive sequence voltage and current at the M end and the positive sequence voltage and current at the N end of the two lines before the fault are collected, and the A in formula (20) is obtained by matrix calculation. n A n-1 …A1e jδ Or B in formula (21) n B n-1 …B1e jδ .
[0081] (2) Substitute the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal after the line fault into equation (16) or equation (23), and we get The real part k1 and imaginary part k2 are used as the line electrical characteristic quantities after data preprocessing.
[0082] (3) A three-layer neural network model is designed, with k1 and k2 as the input of the neural network and the corresponding fault location as the output of the neural network. The present invention comprehensively considers the output error and overfitting phenomenon, selects the number of hidden layer neurons as 10, and the activation function is the tansig function.
[0083] (4) The network is trained using historical fault data to fit the distance measurement function in equation (19) or equation (24). After the training is completed, the functional relationship between the line electrical characteristic quantity and the fault location is stored in the neural network in the form of connection weights and thresholds between neurons.
[0084] (5) When a line fault occurs, the fault data of the line is collected and preprocessed in steps (1) and (2). The obtained line electrical characteristic quantities are input into the trained neural network, and the output of the network is the corresponding fault location.
[0085] The method of the present invention takes into account the problem of uneven distribution of line parameters due to the influence of soil, climate and other factors in various regions on long-distance power transmission lines, as well as the sampling asynchronous angle of data at both ends of the transmission line. The fault ranging function is fitted by a neural network. The ranging process does not rely on the line model and does not require line parameters. In addition, the line fault features are aggregated during the data preprocessing process, so that the neural network training does not require a large number of historical fault samples. The ranging accuracy of the method of the present invention is not affected by factors such as fault type, fault location, transition resistance, fault initial phase angle, load current and asynchronous data.
[0086] Although the embodiments and drawings of the present invention are disclosed for illustrative purposes, those skilled in the art will appreciate that various substitutions, changes and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the contents disclosed in the embodiments and drawings.
Claims
1. A method for asynchronous fault location of two-terminal long-distance transmission lines combining model mechanism and data drive, Features: The steps of the method are: Construct an equivalent model of a two-terminal long-distance transmission line; The equivalent model mechanism analysis of double-terminal long-distance transmission lines is carried out under the distributed parameter model and Π-type line model; Acquire data and use data-driven fault location measurement for double-terminal long-distance transmission lines.
2. The method for non-synchronous fault location of a two-terminal long-distance transmission line combining the model mechanism with data drive according to claim 1, Features: The equivalent model of the two-terminal long-distance transmission line takes into account that the two-terminal long-distance transmission line is affected by soil and climate factors in various regions, resulting in uneven distribution of line parameters. The two-terminal long-distance transmission line is regarded as a non-uniform transmission line connected by n sections of lines with different parameters and lengths. The length of the i-th section is l i , the total length of the transmission line is L n .
3. The method for non-synchronous fault location of a two-terminal long-distance transmission line combining the model mechanism with data drive according to claim 1, Features: Each line segment of the distributed parameter model is equivalent to the distributed parameter model, and the positive sequence wave impedance and positive sequence propagation constant of the i-th line segment are Z ci and γ i ; Each section of the Π-type line model is equivalent to a Π-type line model, and the positive sequence impedance and positive sequence admittance of the i-th section of the line are Z i and Y i .
4. The method for non-synchronous fault location of a two-terminal long-distance transmission line combining the model mechanism with data drive according to claim 1, Features: The analysis of the mechanism of the equivalent model of the double-terminal long-distance transmission line under the distributed parameter model is specifically as follows: When a two-terminal long-distance transmission line operates normally, the relationship between the positive sequence voltage and positive sequence current at both ends of the line is as follows: in: and They are the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal before the line fault respectively; δ is the asynchronous angle of the data collected by the M end relative to the data collected by the N end, taking into account the asynchronous sampling of the data at both ends; If a short circuit fault occurs in a double-terminal long-distance transmission line, the distance measurement function of the double-terminal long-distance transmission line based on the distributed parameter model is derived according to the fault analysis: in: and They are respectively the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal after the line fault.
5. The method for non-synchronous fault location of a two-terminal long-distance transmission line combining the model mechanism with data drive according to claim 1, Features: The analysis of the mechanism of the equivalent model of the double-terminal long-distance transmission line under the Π-type line model is specifically as follows: When a two-terminal long-distance transmission line operates normally, the relationship between the positive sequence voltage and positive sequence current at both ends of the line is: If a double-terminal long-distance transmission line fails, after fault analysis and formula derivation, the distance measurement function of the double-terminal long-distance transmission line based on the Π-type line model can be obtained as follows:
6. The method for non-synchronous fault location of a two-terminal long-distance transmission line combining the model mechanism with data drive according to claim 1, Features: The acquisition of data and the use of data-driven fault location measurement for a two-terminal long-distance power transmission line are specifically as follows: Acquire data, obtain the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal before and after the line fault; The method of using data-driven fault location measurement for a two-terminal long-distance transmission line includes: preprocessing data, designing and training a two-terminal long-distance transmission line fault location measurement model, and performing fault location measurement according to the trained model; The data is preprocessed to collect the positive sequence voltage and current at the M end and the positive sequence voltage and current at the N end before the two sets of line faults, and the A in formula (1) is obtained by matrix calculation. n A n-1 …A 1 e jδ Or B in formula (6) n B n-1 …B 1 e jδ ; Substitute the positive sequence voltage and current at the M terminal and the positive sequence voltage and current at the N terminal after the line fault into equation (4) or equation (9), and we can get The real part k 1 and the imaginary part k 2 As the line electrical characteristic quantity after data preprocessing; Designing and training a two-terminal long-distance transmission line fault location model involves designing a three-layer neural network model with k 1 and k 2 As the input of the neural network, the corresponding fault position is used as the output of the neural network. The present invention comprehensively considers the output error and overfitting phenomenon, selects the number of neurons in the hidden layer as 10, and the activation function is the tansig function; The network is trained using historical fault data to fit the distance measurement function in equation (3) or equation (8). After the training is completed, the functional relationship between the line electrical characteristic quantity and the fault location is stored in the neural network in the form of connection weights and thresholds between neurons. Fault distance measurement based on the trained model includes collecting line fault data for preprocessing when a line fault occurs, inputting the obtained line electrical characteristic quantity into the trained fault distance measurement model, and the output of the model is the corresponding fault location.
