Line fault judgment system and method

Abstract

The invention provides a line fault judgment system and method, and belongs to the field of line fault judgment. The problem of low line fault judgment efficiency is solved; the method specifically comprises the following steps: acquiring historical power consumption data, and calculating expected power consumption; defining different voltage-power change models according to the number of the power supply ends of the power transmission line and the number of the municipal areas, calculating expected receiving power and expected receiving voltage in combination with expected power consumption of each municipal area, and comparing the expected receiving power and the expected receiving voltage with actual receiving power and actual receiving voltage to determine the position of a fault; feeding back the position of the fault on the power transmission line, constructing a duplex AI watchman model, and deploying and continuously monitoring the power transmission line; through obtaining, processing and analyzing the related data of the power transmission line, the change rule of the power and the voltage of each transformer substation on the power transmission line is defined, and the line fault inspection efficiency is improved.

Description

Technical Field

[0001] The present invention provides a line fault determination system and method, and relates to the field of fault determination. Background Art

[0002] Existing methods or systems for determining faults in power transmission lines have the following deficiencies: Single detection methods and indicators: Existing transmission line fault determination systems primarily rely on electrical parameter measurement and signal analysis technologies to identify line anomalies. Commonly used fault detection methods include impedance detection, traveling wave detection, and transient analysis. These methods still face many accuracy challenges in practical applications. The impedance method is easily affected by changes in system operating mode, while the detection accuracy of the traveling wave method is limited by the sensor's installation location and signal attenuation.

[0003] Slow response speed: Existing transmission line fault determination methods often take tens of milliseconds or even longer from the occurrence of a fault to the output of the determination result. This delay mainly comes from the various links of signal acquisition, transmission and processing. Especially in long-distance transmission lines, signal transmission delays and time synchronization issues between different monitoring points will further reduce the system's response speed.

[0004] Complex maintenance: Existing transmission line fault determination systems require the deployment of a large number of high-precision sensors and reliable communication networks, resulting in high initial investment costs. In addition, the system installation process often requires line power outages, which further increases indirect costs. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the present invention aims to provide a line fault determination system and method, aiming to solve the problem of low efficiency in line fault determination.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions: A line fault determination system includes: Data acquisition module: used to obtain the number of urban districts in the target area and obtain the historical electricity consumption data of each urban district; perform time series analysis on the historical electricity consumption data and calculate the expected electricity consumption of each urban district; Data analysis module: This module is used to count the number of power supply terminals, municipal districts, and substations corresponding to each transmission line. Based on the number of power supply terminals and municipal districts, different voltage-power variation models are defined. Combined with the expected power consumption of each municipal district, the expected received power and expected received voltage of each substation on each transmission line are calculated. The actual received power and voltage of each substation on each transmission line are obtained to determine the location of the fault on the transmission line. Line fitting module: used to feedback the location of faults on the transmission line, build a dual AI operator model, and deploy it in the substation corresponding to each transmission line, and continuously monitor the power and voltage changes on each transmission line.

[0007] Furthermore, the specific process of calculating the expected power consumption is as follows: Obtain the hourly electricity consumption of the first municipal district; calculate the expected hourly electricity consumption; Use ADF detection to perform differential detection on the power consumption at each hour in turn to obtain the differential order d; Perform d-order difference on the hourly electricity consumption and process it with autocorrelation function and partial autocorrelation function to obtain autoregressive parameter p and sliding average parameter q; Calculate the autocorrelation coefficient φ （1） ~φ （p） and the sliding average coefficient θ of the 1st to qth order （1） ~θ （q） ; Assume that the expected electricity consumption of the first district in the next day at time t is fe （t） ; Let the autoregressive coefficient of the lth order be φ （l） ; Assume that the electricity consumption of the first district in the past day at t-1 is he （t－l） ; Assume that the sliding average coefficient of the sth order is θ （s） ; Assume that the electricity consumption of the first district in the past day at t-s is he （t－s） ,he （t－s） The corresponding white noise is ε （t－s） ; Construct formula A: ; Based on formula A, perform inverse differential operation to estimate the expected hourly electricity consumption of the first urban district; calculate the expected electricity consumption of the second to mnth urban districts.

[0008] Furthermore, the specific process of defining different voltage-power variation models is as follows: Obtain the number of substations, power supply terminals, and municipal districts on the target line, and define different voltage-power variation models based on the power supply terminals and municipal districts of the target line to determine whether a fault has occurred on the target line; If the target line only has the first municipal district and one power supply terminal, define the voltage-power variation model A: If there is only one power supply terminal on the target line, count the number of parallel jurisdictions sr and define the voltage-power variation model B: If there is only the first municipal district on the target line but multiple power supply terminals, the power supply terminals are considered as parallel districts and the first municipal district is considered as the power supply terminal. Repeat the same process of defining voltage-power variation model B to define voltage-power variation model C: Reverse power flow calculation is performed based on the voltage-power variation model C to determine the fault locations of the target line and the 2nd to mnth corresponding transmission lines.

[0009] Furthermore, the specific process of defining the voltage-power variation model A is as follows: Count the number of substations on the target line (ts+1); Calculate the expected received power Po of the first urban district （1，0） ~Po （1,23） ; Assume that the received power of the first substation is Pe （1） , the receiving voltage is Ue （1） , the transmission power is Pi, and the transmission voltage is Ui; Assume the receiving power of the receiving substation is Po and the receiving voltage is Uo; Assume the branch current is I （1） , I （4） and I （6） , the main circuit current is I （2） , I （3） , I （5） and I （7） ; Let the equivalent impedances be Z and Z （1） , the equivalent admittance is Y （1） 、Y （2） and Y （3） Among them, Y （1） and Y （2） The values of are equal; Definition formula A-1-1: ; Formula A-1-2: ; Combining these formulas, we get formula A-1-3: ; Definition formula A-2-1: ; Formula A-2-2: ; Combining these equations, we get formula A-3: ; Definition formula A-4-1: ;Formula A-4-2: ; The combined formula A-4-3 is: ; Define formula A-5; ; Combining these equations, we get formula A-6: ; Formula A-3 and Formula A-6 are used as voltage-power variation model A; Obtain the rated operating voltage Uw of the receiving substation; According to Uw and Po （1，0） 、Po （1，1） ~Po （1,23） , calculate the expected received power Pe of the 1st to mnth substations at each hour （1，0） ~Pe （tr,23） ;Expected receiving voltage Ue （1，0） ~Ue （tr,23） ; Expected transmission power Pi （1，0） ~Pi （tr,23） ; Expected transmission voltage Ui （1，0） ~Ui （tr,23） ; Define the conditions for determining line faults.

[0010] Furthermore, the specific process of defining the judgment conditions of line faults is as follows: Assume that the actual received power at the xth substation at time y is fPe （x，y） , the actual receiving voltage is fUe （x，y） , the actual transmission power is fPi （x，y） , the actual transmission voltage is fUi （x，y） ; Assume that the expected received power at the xth substation at time y is Pe （x，y） , the expected receiving voltage is Ue （x，y） , the expected transmission power is Pi （x，y） , the expected transmission voltage is Ui （x，y） ; Determining fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） Are the differences within ζ? ζ represents the coefficient of determination. If fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） If the differences are within ζ, then fPo （x，y） With Po （x，y） The difference and fUo （x，y） With Uo （x，y） Are the differences within ζ? If they are all within ζ, the substation is normal; If it is not within ζ, then the substation downstream is abnormal; If fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） If the difference is not within ζ, then determine whether formula A-7 is valid; Formula A-7: ; If formula A-7 holds true, the upstream of the substation is abnormal and the downstream is normal; If formula A-7 holds true, both the upstream and downstream sides of the substation are abnormal; Obtain the actual received power, actual received voltage, actual transmission power, and actual transmission voltage of the 1st to trth substations at each hour, and find the fault area based on the judgment conditions.

[0011] Furthermore, the specific process of defining the voltage-power variation model B is as follows: Get the length of the transmission line from the diversion node to the first municipal district, Li; get the length of the transmission line from the diversion node to the first to srth parallel districts, Lo （1） ~Lo （sr） ; Obtain the current frequency fz on the target line, the conductor spacing Df, conductor radius Dr, conductor relative permeability ur, conductor conductivity σ, air dielectric constant δr, and vacuum dielectric constant δo of the transmission cable; Assume that the length of the transmission line from the diversion node to the vth parallel jurisdiction is Lo （v） , the equivalent resistance Rxe of the vth parallel region （v） ; Define formula B-1: ; Calculate the equivalent resistance Rxe of the srth to srth parallel regions （1） ~Rxe （sr） ; Let the equivalent inductance of v parallel jurisdictions be Lxe （v） , define formula B-2: ; Calculate the equivalent inductance of the 1st to srth parallel jurisdictions; Assume that the equivalent impedance of the vth parallel jurisdiction is Zxe （v） , define formula B-3: ; ij represents the imaginary unit; ω represents the angular frequency of the current ; Let the equivalent admittance of the vth parallel jurisdiction be Yxe （v）, define formula B-4: ; Calculate the equivalent impedance of the 1st to srth parallel jurisdictions.

[0012] Furthermore, the workflow of defining the voltage-power variation model B further includes: Based on Li, calculate the equivalent impedance Zx of the first municipal district; Get the length Ll of the transmission line from the first municipal district to the first parallel municipal district （o-1） , the length of the transmission line from the (sr-1)th parallel city district to the srth parallel city district is Ll （（sr-1）-sr） ; Calculate Ll （o-1） ~Ll （（sr-1）-sr） The corresponding equivalent impedance Zxl （o-1） ~Zxl （（sr-1）-sr） ; Assume that the receiving power of the receiving substation corresponding to the first municipal district is Poo; the receiving power of the receiving substation corresponding to the first to srth parallel districts is Pox （1） ~Pox （sr） ; Assume that the transmission power of the shunt node corresponding to the power station is Ppx, and define formula B-5: ; Pox （v） represents the receiving power of the receiving substation corresponding to the vth parallel jurisdiction; Pox （n） Indicates the receiving power of the receiving substation corresponding to the nth parallel jurisdiction; Zxl （n-（n+1）） It represents the equivalent impedance from the nth parallel city district to the (n+1)th parallel city district; Qex represents the repeated term: ; Formula B-5 is used as voltage-power variation model B, and voltage-power variation model B is used as the basis for reverse power flow calculation in process B22 to determine the fault location of the target line.

[0013] Furthermore, the process of building a dual AI operator model is as follows: Obtain the fault information of each transmission line corresponding to the power station or line, and generate a structured event cluster as the sample set A; The BERT+Transformer+MLP architecture is used as the framework for the dual AI attendant model; Divide sample set A into a training set and a validation set at a ratio of 7:3. Optimize the default parameters of the dual AI attendant model through supervised learning until the model can accurately identify the event cluster and fault type corresponding to the fault information. Obtain misjudgment information about the corresponding power station or line for each transmission line; deploy two independent modes, working and training, in the dual AI operator model; when the model performs line fault inspection, it operates in working mode; when performing model training, it operates in training mode; The misjudgment information is clustered and added to sample set A to obtain sample set B. Sample sets A and B are used alternately to train the AI attendant model again until the model outputs all misjudgment information and the corresponding event clusters and fault types without bias, thus achieving adaptive fault diagnosis without relying on a rule base. The dual AI operator model is deployed in the substation corresponding to each transmission line.

[0014] A line fault determination method includes: Obtain the number of urban districts in the target area and obtain historical electricity consumption data for each urban district; perform time series analysis on the historical electricity consumption data and calculate the expected electricity consumption for each urban district; Count the number of power supply terminals, municipal districts, and substations corresponding to each transmission line. Define different voltage-power variation models based on the number of power supply terminals and municipal districts of the transmission line. Combined with the expected power consumption of each municipal district, calculate the expected received power and expected received voltage of each substation on each transmission line. Obtain the actual received power and actual received voltage of each substation on each transmission line to determine the location of the fault on the transmission line. Feedback the location of the fault on the transmission line, build a dual AI operator model, and deploy it in the substation corresponding to each transmission line, and continuously monitor the power and voltage changes on each transmission line.

[0015] Compared with the prior art, the present invention has the following beneficial effects: Multi-model judgment: The data analysis module of the present invention constructs a variety of voltage-power change models for different power grid topologies. The present invention can intelligently select the most suitable analysis model based on different combinations of the number of power supply terminals and the number of urban districts, thereby achieving highly accurate fault judgment and scenario adaptability.

[0016] Intelligent self-optimization features: This invention adopts the BERT+Transformer+MLP architecture, combined with supervised learning and continuous optimization mechanisms. The system generates structured event clusters from fault information as sample set A, divides the training set and validation set into a 7:3 ratio, and ensures that the model can judge various types of faults without bias through iterative optimization. At the same time, the invention also collects manually corrected misjudgment information, forms sample set B after clustering processing, and uses the two groups of samples alternately for secondary training, thereby improving the long-term applicability and maintenance convenience of the system.

[0017] High Accuracy: This invention combines the dual advantages of physical models and machine learning. Precise modeling based on circuit theory ensures the reliability of basic judgments, while the self-learning ability of the AI algorithm continuously optimizes judgment accuracy, enabling the system to adapt to various boundary conditions and special scenarios. This makes the system suitable for both large regional power grids and small dedicated networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings: Figure 1 Schematic diagram of the system of the present invention; Figure 2 Schematic diagram of the method of the present invention; Figure 3 This is a one-to-one relationship diagram of the present invention; Figure 4 This is a schematic diagram of a one-to-many relationship of the present invention. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1 See also Figure 1 , a line fault determination system comprising: Data acquisition module: used to obtain the number of urban districts in the target area and obtain the historical electricity consumption data of each urban district (yesterday); perform time series analysis on the historical electricity consumption data and calculate the expected electricity consumption of each urban district; It should be noted that the "line" in the present invention refers to a power transmission line; the "target area" in the present invention refers to a city-level area where the present invention (a line fault determination system and method) is used to determine line faults; Process A: The workflow of the data acquisition module is as follows: Get the number of urban districts in the target area, mn; Get the electricity consumption of the first district from 0:00, 1:00, to 23:00 yesterday（1,0） 、he （1,1） ~he （1,23） ; to him （1,0） ~he （1,23） Perform time series analysis to calculate the expected power consumption qr of the first urban district from 0:00, 1:00, to 23:00 in the next day （1,0） 、qr （1,1） ~qr （1,23） ; Use ADF detection to detect he （1,0） ~he （1,23） Perform differential detection to obtain the differential order d; to him （1,0） ~he （1,23） Perform d-order differences and process them with the autocorrelation function (ACF) and partial autocorrelation function (PACF) to obtain the autoregressive parameter p and the moving average parameter q (the range of d, p, and q is [1, 24), and the values of (24-p) and (24-q) are both greater than or equal to 1); Calculate the autocorrelation coefficient φ of the first, second, and pth order （1） 、φ （2） ~φ （p） and the sliding average coefficient θ of the 1st to qth order （1） ,θ （2） ~θ （q） ; According to he （1,0） 、he （1,1） ~he （1,23） Calculate the autocorrelation coefficient φ （1，1） 、φ （1，2） ~φ （1，p） and the sliding average coefficient θ of the 1st to qth order （1，1） ,θ （1，2） ~θ （1，q） ; Calculate he （1,0） ~he （1,23） The average value ahe; will {he （1,0） ~he （1,23）} as sequence H; Calculate the 1st to pth order difference corresponding to sequence H and obtain sequence H （1） ~H （p） ; Calculate the sequence H with respect to the sequence H （1） ~H （p） The autocovariance of γ （1） ~γ （p） ; Construct a (1×p) matrix B （φ） : ; Construct a (1×p) matrix B （1） : ; Construct a (p×p) matrix B （2） : ; Among them, the matrix B （2） The elements on the diagonal are all ahe; the sequence H is about the sequence H （1） ~H （p-1） The autocovariance γ （1） ~γ （p-1） , symmetrically arranged in matrix B （2） On both sides of the diagonal line; Calculate φ （1，1） ~φ （1，p） Value: ; Among them, -1 represents the inverse of the matrix; Compare the size of p and q, and calculate the sliding average coefficient θ from the 1st to the qth order （1） ~θ （q） ; If p≥q, then according to the sequence H about the sequence H （1） ~H （q） The autocovariance of γ （1） ~γ （q） , construct the (1×q) matrix Bφ （q） : ; Construct a (1×q) matrix B （θ） : ; Construct a (q×q) matrix Bq: ; Among them, the elements on the diagonal of matrix Bq are all sod; the sequence H is about the sequence H （1） ~H （q-1） The autocovariance γ （1） ~γ （q-1） , symmetrically arranged in matrix B （2） On both sides of the diagonal line; Calculate θ （1） ~θ （q） Value: ; If p < q, calculate the 1st to qth order difference of sequence H and obtain sequence Hq （1） ~Hq （q） ; Calculate the sequence H with respect to the sequence Hq （1） ~Hq （q） The autocovariance of γq is obtained （1） ~γq （q） ; Define the function ss(i): ; Among them, γq（j） Denotes sequence H with respect to sequence Hq （j） The autocovariance of θ （i） ,θ （j） and θ （j+i） , respectively represent the sliding average coefficients of the i-th order, the i-th order, and the (j+i)-th order, and the value range of i and j is: 1 ~ (q-1); Construct the matrix equation: ; where θ （j+1） represents the sliding mean coefficient of the (j+1)th order; γq （1） ~γq （q） Substitute into the matrix equation and calculate θ （1） ~θ （q） The value of Calculate he （1,0） 、he （1,1） ~he （1,23） The average value Hhe; Calculate he （1,0） ~he （1,23） White noise ε （0） ~ε （23） ; Among them, ε （0） =he （1,0） -Hhe;ε （1） =he （1,1） -Hhe; and so on, ε （23） =he （1,23） -Hhe; Assume that the expected electricity consumption of the first district in the next day at time t is fe （t） ; where t≤d; Assume that the autoregressive coefficient of the lth order is φ （l） ; Assume that the electricity consumption of the first district in the past day (t-l) is he （t－l） ; Among them, the value range of l is: 1~p; Assume that the sliding mean coefficient of the sth order is θ （s） ; Assume that the electricity consumption of the first district in the past day (t-s) is he （t－s） ,he （t－s） The corresponding white noise is ε （t－s） ; Among them, the value range of s is: 1~q; ε （t－s） ∈{ε （0） ~ε （23）}; Construct formula A: ; Based on formula A, perform inverse differential calculation to estimate the expected power consumption qr of the first urban district at 0:00, 1:00, and until 23:00 in the next day. （1，0） 、qr （1，1） ~qr （1,23） ; Repeated calculation of qr （1，0） ~qr （1,23） The same steps are used to calculate the expected electricity consumption qr of the 2nd to mnth districts. （2，0） ~qr （mn,23） .

[0021] Data analysis module: This module is used to count the number of power supply terminals, municipal districts, and substations corresponding to each transmission line. Based on the number of power supply terminals (i.e., power stations or other municipal or provincial power grids) and the number of municipal districts, different voltage-power variation models are defined. Combined with the expected power consumption of each municipal district, the expected received power and voltage at each substation on each transmission line are calculated. The actual received power and voltage at each substation on each transmission line are obtained to determine the location of the fault on the transmission line. Process B: The specific process of the data analysis module is as follows: Process B1: The transmission line supplying power to the first municipal district is selected as the target line; Obtain the number of substations, power supply terminals, and municipal districts on the target line, and define different voltage-power variation models based on the power supply terminals and municipal districts of the target line to determine whether a fault has occurred on the target line; Process B2: Please refer to Figure 3 If there is only the first municipal district and one power supply terminal on the target line, that is, a "one-to-one" relationship, then the voltage-power change model A is defined as: Count the number of substations on the target line (ts+1). The substation directly supplying power to the first municipal district is considered the receiving substation. With the receiving substation as the starting point, there are a total of ts substations on the target line. Get the expected electricity consumption qr of the first district in the next day at 0:00, 1:00, and until 23:00 in the first district （1，0） 、qr （1，1） ~qr （1,23） ; Calculate the expected receiving power Po of the receiving substation in the first municipal district from 0:00, 1:00 to 23:00 on the next day （1，0） 、Po （1，1） ~Po （1,23） ; Among them, Po （1，0） =qr （1，0） / (60×60);Po （1，1） =qr（1，1） / (60 × 60); and so on, Po （1，23） =qr （1，23） / (60×60); Process B21: Assume that the received power of the first substation is Pe （1） , the receiving voltage is Ue （1） , the transmission power is Pi, and the transmission voltage is Ui; Assume the receiving power of the receiving substation is Po and the receiving voltage is Uo; Assume the branch current is I （1） , I （4） and I （6） , the main circuit current is I （2） , I （3） , I （5） and I （7） ; Let the equivalent impedances be Z and Z （1） , the equivalent admittance is Y （1） 、Y （2） and Y （3） Among them, Y （1） and Y （2） The values of are equal; Define voltage-power variation model A: According to Kirchhoff's current law, we know that: (Formula A-1-1); According to Ohm's law, we know that: (Formula A-1-2); Combining formula A-1-1 and formula A-1-1, we get formula A-1-3: (Formula A-1-3); According to Joule's law, we know that: (Formula A-2-1); (Formula A-2-2); Formula A-2-2 is transformed into: (Formula A-2-3); Substituting Formula A-1-3 and Formula A-2-3 into Formula A-2-2, we obtain Formula A-2-4: (Formula A-2-4); Combining Formula A-1-1, Formula A-1-2, and Formula A-2-4, we get Formula A-3: (Derivation process of Formula A-3); From the derivation process of formula A-3, we can get formula A-3: (Formula A-3); According to Kirchhoff's current law, we know that: (Formula A-4-1); According to Joule's law, we know that: (Formula A-4-2); Combining formula A-1-2 and formula A-4-2, we get formula A-4-3: (Formula A-4-3); According to Ohm's law, we know that; (Formula A-5-1); Combining Formula A-4-2 and Formula A-5-1, and transforming them, we get: (Formula A-5-2); Combining Formula A-4-3 and Formula A-5-2, we get Formula A-6: (Formula A-6); Formula A-3 and Formula A-6 are used as voltage-power variation model A; Get the rated working voltage of the receiving substation, denoted as Uw; Process B22: According to the voltage-power variation model A, take Uw as Uo and Po as （1，0） 、Po （1，1） ~Po （1,23） For Po, use reverse power flow calculation to calculate the expected receiving power Pe of the first substation in the first urban district at 0:00, 1:00 and until 23:00 in the next day. （1，0） 、Pe （1，1） ~Pe （1,23） ;Expected receiving voltage Ue （1，0） 、Ue （1，1） ~Ue （1,23） ; Expected transmission power Pi （1，0） 、Pi （1，1） ~Pi （1,23） ; Expected transmission voltage Ui （1，0） 、Ui （1，1） ~Ui （1,23） ; The expected received power Pe of the first urban district at the second substation at 0:00, 1:00 and until 23:00 in the next day （2，0） 、Pe （2，1） ~Pe （2,23） ;Expected receiving voltage Ue （2，0） 、Ue（2，1） ~Ue （2,23） ; Expected transmission power Pi （2，0） 、Pi （2，1） ~Pi （2,23） ; Expected transmission voltage Ui （2，0） 、Ui （2，1） ~Ui （2,23） ; Similarly, the expected received power Pe of the first city district at the tr-th substation at 0:00, 1:00, and until 23:00 on the next day is （tr，0） 、Pe （tr，1） ~Pe （tr,23） ;Expected receiving voltage Ue （tr，0） 、Ue （tr，1） ~Ue （tr,23） ; Expected transmission power Pi （tr，0） 、Pi （tr，1） ~Pi （tr,23） ; Expected transmission voltage Ui （tr，0） 、Ui （tr，1） ~Ui （tr,23） ; Define the conditions for determining line faults; Process B221: Assume that the actual received power of the xth substation at time y on the next day is fPe （x，y） , the actual receiving voltage is fUe （x，y） , the actual transmission power is fPi （x，y） , the actual transmission voltage is fUi （x，y） ; Assume that the expected received power of the xth substation at time y in the future day is Pe （x，y） , the expected receiving voltage is Ue （x，y） , the expected transmission power is Pi （x，y） , the expected transmission voltage is Ui （x，y） ; Among them, the value range of x is: 1~tr; the value range of y is: 0~23; Determining fPe （x，y） With Pe （x，y） The (numerical) difference and fUe （x，y） With Ue （x，y） The (numerical) differences are all within ζ; where ζ represents the determination coefficient; the value of ζ is 0.01; users or relevant technical personnel can adjust the value of ζ according to actual needs; Process B222: If fPe （x，y） With Pe （x，y） The (numerical) difference and fUe （x，y） With Ue （x，y） If the (numerical) differences are within ζ, then fPo is determined. （x，y） With Po（x，y） The (numerical) difference and fUo （x，y） With Uo （x，y） Are the (numerical) differences within ζ? If fPo （x，y） With Po （x，y） The (numerical) difference and fUo （x，y） With Uo （x，y） If the (value) differences are all within ζ, it means that the x-th substation is normal; If fPo （x，y） With Po （x，y） The (numerical) difference and fUo （x，y） With Uo （x，y） If the difference in values is not within ζ, it means that an abnormality occurs downstream of the x-th substation; Process B223: If fPe （x，y） With Pe （x，y） The (numerical) difference and fUe （x，y） With Ue （x，y） If the difference in values is not within ζ, then determine whether Formula A-7 is valid. Formula A-7: ; If formula A-7 holds true, it means that the upstream of the x-th substation is abnormal, but the downstream is normal; If formula A-7 holds true, it means that abnormalities have occurred both upstream and downstream of the x-th substation; Process B224: According to the judgment conditions, the actual received power, actual received voltage, actual transmission power and actual transmission voltage of the 1st to trth substations from 0:00 to 23:00 on the next day are obtained, and the corresponding expected received power, expected received voltage, expected transmission power and expected transmission voltage (i.e., Pe （1，0） ~Pe （tr,23） , Ue （1，0） ~Ue （tr,23） , Pi （1，0） ~Pi （tr,23） and Ui （1，0） ~Ui （tr,23） ) Compare and find out the fault area; Process B3: If the target line has only one power supply terminal but multiple municipal districts, that is, a "one-to-many" relationship, define voltage-power variation model B: The other municipal districts on the target route except the first municipal district are regarded as parallel districts; the number of parallel districts sr is counted; See also Figure 4 , define voltage-power variation model B; Process B31: Get the length of the transmission line from the diversion node to the first city district, and get Li; get the length of the transmission line from the diversion node to the first, second, and so on to the srth parallel districts, and get Lo （1） 、Lo （2） ~Lo （sr） ; Obtain the current frequency fz on the target line, the conductor spacing Df, conductor radius Dr, conductor relative permeability ur, conductor conductivity σ, air dielectric constant δr, and vacuum dielectric constant δo of the transmission cable (target line); Assume that the length of the transmission line from the diversion node to the vth parallel jurisdiction is Lo （v） , the equivalent resistance Rxe of the vth parallel region （v） ; The value range of v is: 1~sr; Define formula B-1: ; According to formula B-1, calculate the equivalent resistance Rxe of the 1st, 2nd, and srth parallel regions. （1） 、Rxe （2） ~Rxe （sr） ; Let the equivalent inductance of v parallel jurisdictions be Lxe （v） , define formula B-2: ; According to formula B-1, calculate the equivalent inductance Lxe of the 1st, 2nd, and srth parallel jurisdictions （1） 、Lxe （2） ~Lxe （sr） ; Let the equivalent capacitance of v parallel regions be Cxe （v） , define formula B-3: ; According to formula B-3, calculate the equivalent capacitance Cxe of the 1st, 2nd, and srth parallel regions. （1） , Cxe （2） ~Cxe （sr） ; Let the equivalent conductance of v parallel regions be Gxe （v） , define formula B-4: ; According to formula B-4, calculate the equivalent conductance Gxe of the 1st, 2nd, and srth parallel regions. （1） 、Gxe （2） ~Gxe （sr） ; Assume that the equivalent impedance of the vth parallel jurisdiction is Zxe （v）, define formula B-5: ; where ij represents the imaginary unit; ω represents the angular frequency of the current ; Let the equivalent admittance of the vth parallel jurisdiction be Yxe （v） , define formula B-6: ; Among them, ij represents the imaginary unit; According to formula B-5 and formula B-6, calculate the equivalent impedance Zxe of the 1st, 2nd, and srth parallel jurisdictions （1） 、Zxe （2） ~Zxe （sr） ; Equivalent admittance Yxe （1） 、Yxe （2） ~Yxe （sr） ; Replace Lo in Formula B-1 to Formula B-4 （v） Replace Li and calculate the equivalent resistance, equivalent inductance, equivalent capacitance, and equivalent susceptance of the first district. Then, according to Formulas B-5 to B-6, calculate the equivalent impedance Zx and equivalent admittance Yx of the first district. Define the power flow formula corresponding to voltage-power variation model B, and obtain formula B-7; Get the length Ll of the transmission line from the first municipal district to the first parallel municipal district （o-1） , the length of the transmission line from the first parallel city district to the second parallel city district Ll （1-2） , the length of the transmission line from the second parallel city district to the third parallel city district Ll （2-3） ; Similarly, the length of the transmission line from the (sr-1)th parallel city district to the srth parallel city district is Ll （（sr-1）-sr） ; Calculate Ll according to formula B-1 to formula B-6 （o-1） , Ll （1-2） , Ll （2-3） ~Ll （（sr-1）-sr） The corresponding equivalent impedance Zxl （o-1） 、Zxl （1-2） 、Zxl （2-3） ~Zxl （（sr-1）-sr） ; Assume that the receiving power of the first municipal district corresponding to the receiving substation is Poo; the receiving power of the first, second, and up to the srth parallel districts corresponding to the receiving substation is Pox （1） , Pox （2） ~Pox （sr） ; Assume that the transmission power of the shunt node corresponding to the power station is Ppx, and define formula B-7: ; Pox （v） represents the receiving power of the receiving substation corresponding to the vth parallel jurisdiction, Pox （n） Indicates the receiving power of the receiving substation corresponding to the nth parallel jurisdiction, and the value range of n is: 1 to (sr-1); Zxl （n-（n+1）） It represents the equivalent impedance from the nth parallel city district to the (n+1)th parallel city district; Qex represents the repeated item, and the calculation formula of Qex is: ; Use Equation B-7 as voltage-power variation model B, and use voltage-power variation model B as the basis for reverse power flow calculation in process B22 to determine the fault location of the target line. Process B4: If there is only the first municipal district on the target line, but there are multiple power supply terminals; that is, a "many-to-one" relationship, then the power supply terminals are considered as parallel districts, and the first municipal district is considered as the power supply terminal. Repeat the same process as process B3 to define voltage-power variation model B to define voltage-power variation model C: Perform reverse power flow calculation based on voltage-power variation model C to determine the fault location of the target line; Repeat the same process of determining whether the transmission line corresponding to the first municipal district is faulty, and determine whether the second to mnth corresponding transmission lines are faulty.

[0022] Line fitting module: This module provides feedback on the location of faults on transmission lines and builds a dual "AI operator" model. This dual "AI operator" model is deployed in the substation corresponding to each transmission line and continuously monitors power and voltage changes on each transmission line. The process of building a dual "AI attendant" model is as follows: Obtain the fault information of each transmission line corresponding to the power station or line, and generate a structured event cluster as the sample set A; The BERT+Transformer+MLP architecture is used as the framework for the dual "AI Attendant" model; Split sample set A into a training set and a validation set in a 7:3 ratio. Optimize the default parameters of the dual "AI Attendant" model (i.e., the default parameters in the BERT+Transformer+MLP architecture) through supervised learning until the "AI Attendant" model can accurately determine the event cluster and fault type corresponding to fault information. Obtain misjudgment information of the substation or line corresponding to each transmission line (i.e., erroneous fault information corrected by substation staff); In the dual "AI Operator" model, two independent modes, "working" and "training," are deployed. When the model is performing line fault inspection, it operates in "working" mode; when performing model training, it operates in "training" mode. The misjudgment information is clustered and added to sample set A to obtain sample set B. Sample sets A and B are used alternately to train the "AI Operator" model again until the model can output all misjudgment information and the corresponding event clusters and fault types without bias. This forms a closed-loop system of "signal acquisition-event cluster generation-intelligent judgment-feedback optimization", realizing adaptive fault diagnosis functions that do not rely on a rule base. The dual "AI operator" model is deployed in the substation corresponding to each transmission line.

[0023] Example 2 See also Figure 2 , a line fault determination method includes: Step S1: Obtain the number of urban districts in the target area and obtain the historical electricity consumption data (yesterday) of each urban district; perform time series analysis on the historical electricity consumption data and calculate the expected electricity consumption of each urban district; Step S2: Count the number of power supply terminals, the number of urban districts, and the number of substations corresponding to each transmission line; define different voltage-power variation models based on the number of power supply terminals and the number of urban districts of the transmission line, and calculate the expected received power and expected received voltage of each substation on each transmission line based on the expected power consumption of each urban district; obtain the actual received power and actual received voltage of each substation on each transmission line to determine the location of the fault on the transmission line; Step S3: Feedback the location of the fault on the transmission line and build a dual "AI operator" model; deploy the dual "AI operator" model in the substation corresponding to each transmission line and continuously monitor the power and voltage changes on each transmission line.

[0024] The above formulas are all dimensionless and calculated by taking their numerical values. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions. For example, if there are weight coefficients and proportional coefficients, the size of the settings is to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. Regarding the size of the weight coefficient and the proportional coefficient, as long as it does not affect the proportional relationship between the parameter and the quantized value, it is fine.

[0025] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed by the present invention, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A line fault determination system, characterized in that: The system comprises: Data acquisition module: used to obtain the number of urban districts in the target area and obtain the historical electricity consumption data of each urban district; perform time series analysis on the historical electricity consumption data and calculate the expected electricity consumption of each urban district; Data analysis module: This module is used to count the number of power supply terminals, municipal districts, and substations corresponding to each transmission line. Based on the number of power supply terminals and municipal districts, different voltage-power variation models are defined. Combined with the expected power consumption of each municipal district, the expected received power and expected received voltage of each substation on each transmission line are calculated. The actual received power and voltage of each substation on each transmission line are obtained to determine the location of the fault on the transmission line. Line fitting module: used to feedback the location of faults on the transmission line, build a dual AI operator model, and deploy it in the substation corresponding to each transmission line, and continuously monitor the power and voltage changes on each transmission line.

2. A line fault determination system according to claim 1, characterized in that: The specific process of calculating expected power consumption is as follows: Obtain the hourly electricity consumption of the first municipal district; calculate the expected hourly electricity consumption; Use ADF detection to perform differential detection on the power consumption at each hour in turn to obtain the differential order d; Perform d-order difference on the hourly electricity consumption and process it with autocorrelation function and partial autocorrelation function to obtain autoregressive parameter p and sliding average parameter q; Calculate the autocorrelation coefficient φ （1） ~φ （p） and the sliding average coefficient θ of the 1st to qth order （1） ~θ （q） ; Assume that the expected electricity consumption of the first district in the next day at time t is fe （t） ; Let the autoregressive coefficient of the lth order be φ （l） ; Assume that the electricity consumption of the first district in the past day at t-1 is he （t－l） ; Assume that the sliding mean coefficient of the sth order is θ （s） ; Assume that the electricity consumption of the first district in the past day at t-s is he （t－s） ,he （t－s） The corresponding white noise is ε （t－s） ; Construct formula A: ； Based on formula A, perform inverse differential operation to estimate the expected hourly electricity consumption of the first urban district; calculate the expected electricity consumption of the second to mnth urban districts.

3. A line fault determination system according to claim 1, characterized in that: The specific process of defining different voltage-power variation models is as follows: Obtain the number of substations, power supply terminals, and municipal districts on the target line, and define different voltage-power variation models based on the power supply terminals and municipal districts of the target line to determine whether a fault has occurred on the target line; If the target line only has the first municipal district and one power supply terminal, define the voltage-power variation model A: If there is only one power supply terminal on the target line, count the number of parallel jurisdictions sr and define the voltage-power variation model B: If there is only the first municipal district on the target line but multiple power supply terminals, the power supply terminals are considered as parallel districts and the first municipal district is considered as the power supply terminal. Repeat the same process of defining voltage-power variation model B to define voltage-power variation model C: Reverse power flow calculation is performed based on the voltage-power variation model C to determine the fault locations of the target line and the 2nd to mnth corresponding transmission lines.

4. A line fault determination system according to claim 3, characterized in that: The specific process of defining the voltage-power variation model A is as follows: Count the number of substations on the target line (ts+1); Calculate the expected received power Po of the first urban district （1，0） ~Po （1,23） ; Assume that the received power of the first substation is Pe （1） , the receiving voltage is Ue （1） , the transmission power is Pi, and the transmission voltage is Ui; Assume the receiving power of the receiving substation is Po and the receiving voltage is Uo; Assume the branch current is I （1） , I （4） and I （6） , the main circuit current is I （2） , I （3） , I （5） and I （7） ; Let the equivalent impedances be Z and Z （1） , the equivalent admittance is Y （1） 、Y （2） and Y （3） Among them, Y （1） and Y （2） The values of are equal; Definition formula A-1-1: ; Formula A-1-2: ; Combining these formulas, we get formula A-1-3: ； Definition formula A-2-1: ; Formula A-2-2: ; Combining these equations, we get formula A-3: ； Definition formula A-4-1: ;Formula A-4-2: ; The combined formula A-4-3 is: ; Define formula A-5; ； Combining these equations, we get formula A-6: ； Formula A-3 and Formula A-6 are used as voltage-power variation model A; Obtain the rated operating voltage Uw of the receiving substation; According to Uw and Po （1，0） 、Po （1，1） ~Po （1,23） , calculate the expected received power Pe of the 1st to mnth substations at each hour （1，0） ~Pe （tr,23） ;Expected receiving voltage Ue （1，0） ~Ue （tr,23） ; Expected transmission power Pi （1，0） ~Pi （tr,23） ; Expected transmission voltage Ui （1，0） ~Ui （tr,23） ; Define the conditions for determining line faults.

5. A line fault determination system according to claim 4, characterized in that: The specific process of defining the judgment conditions for line faults is as follows: Assume that the actual received power at the xth substation at time y is fPe （x，y） , the actual receiving voltage is fUe （x，y） , the actual transmission power is fPi （x，y） , the actual transmission voltage is fUi （x，y） ; Assume that the expected received power at the xth substation at time y is Pe （x，y） , the expected receiving voltage is Ue （x，y） , the expected transmission power is Pi （x，y） , the expected transmission voltage is Ui （x，y） ; Determining fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） Are the differences within ζ? ζ represents the coefficient of determination. If fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） If the differences are within ζ, then fPo （x，y） With Po （x，y） The difference and fUo （x，y） With Uo （x，y） Are the differences within ζ? If they are all within ζ, the substation is normal; If it is not within ζ, then the substation downstream is abnormal; If fPe （x，y） With Pe （x，y） The difference and fUe （x，y） With Ue （x，y） If the difference is not within ζ, then determine whether formula A-7 is valid; Formula A-7: ； If formula A-7 holds true, the upstream of the substation is abnormal and the downstream is normal; If formula A-7 holds true, both the upstream and downstream sides of the substation are abnormal; Obtain the actual received power, actual received voltage, actual transmission power, and actual transmission voltage of the 1st to trth substations at each hour, and find the fault area based on the judgment conditions.

6. A line fault determination system according to claim 3, characterized in that: The specific process of defining the voltage-power variation model B is as follows: Get the length of the transmission line from the diversion node to the first municipal district, Li; get the length of the transmission line from the diversion node to the first to srth parallel districts, Lo （1） ~Lo （sr） ; Obtain the current frequency fz on the target line, the conductor spacing Df, conductor radius Dr, conductor relative permeability ur, conductor conductivity σ, air dielectric constant δr, and vacuum dielectric constant δo of the transmission cable; Assume that the length of the transmission line from the diversion node to the vth parallel jurisdiction is Lo （v） , the equivalent resistance Rxe of the vth parallel region （v） ; Define formula B-1: ； Calculate the equivalent resistance Rxe of the srth to srth parallel regions （1） ~Rxe （sr） ; Let the equivalent inductance of v parallel jurisdictions be Lxe （v） , define formula B-2: ； Calculate the equivalent inductance of the 1st to srth parallel jurisdictions; Assume that the equivalent impedance of the vth parallel jurisdiction is Zxe （v） , define formula B-3: ; ij represents the imaginary unit; ω represents the angular frequency of the current ; Let the equivalent admittance of the vth parallel jurisdiction be Yxe （v） , define formula B-4: ； Calculate the equivalent impedance of the 1st to srth parallel jurisdictions.

7. A line fault determination system according to claim 6, characterized in that: The workflow of defining the voltage-power variation model B further includes: Based on Li, calculate the equivalent impedance Zx of the first municipal district; Get the length Ll of the transmission line from the first municipal district to the first parallel municipal district （o-1） , the length of the transmission line from the (sr-1)th parallel city district to the srth parallel city district is Ll （（sr-1）-sr） ; Calculate Ll （o-1） ~Ll （（sr-1）-sr） The corresponding equivalent impedance Zxl （o-1） ~Zxl （（sr-1）-sr） ; Assume that the receiving power of the receiving substation corresponding to the first municipal district is Poo; the receiving power of the receiving substation corresponding to the first to srth parallel districts is Pox （1） ~Pox （sr） ; Assume that the transmission power of the shunt node corresponding to the power station is Ppx, and define formula B-5: ； Pox （v） represents the receiving power of the receiving substation corresponding to the vth parallel jurisdiction; Pox （n） Indicates the receiving power of the receiving substation corresponding to the nth parallel jurisdiction; Zxl （n-（n+1）） It represents the equivalent impedance from the nth parallel city district to the (n+1)th parallel city district; Qex represents the repeated term: ； Formula B-5 is used as voltage-power variation model B, and voltage-power variation model B is used as the basis for reverse power flow calculation in process B22 to determine the fault location of the target line.

8. A line fault determination system according to claim 1, characterized in that: The process of building a dual AI operator model is as follows: Obtain the fault information of each transmission line corresponding to the power station or line, and generate a structured event cluster as the sample set A; The BERT+Transformer+MLP architecture is used as the framework for the dual AI attendant model; Divide sample set A into a training set and a validation set at a ratio of 7:

3. Optimize the default parameters of the dual AI attendant model through supervised learning until the model can accurately identify the event cluster and fault type corresponding to the fault information. Obtain misjudgment information on the corresponding power station or line for each transmission line; deploy two independent working and training modes in the dual AI operator model; When the model performs line fault detection, it runs in working mode; when it performs model training, it runs in training mode; The misjudgment information is clustered and added to sample set A to obtain sample set B. Sample sets A and B are used alternately to train the AI attendant model again until the model outputs all misjudgment information and the corresponding event clusters and fault types without bias, thus achieving adaptive fault diagnosis without relying on a rule base. The dual AI operator model is deployed in the substation corresponding to each transmission line.

9. A line fault determination method, applicable to a line fault determination system according to any one of claims 1 to 8, characterized in that: The determination method includes: Obtain the number of urban districts in the target area and obtain historical electricity consumption data for each urban district; perform time series analysis on the historical electricity consumption data to calculate the expected electricity consumption for each urban district; Count the number of power supply terminals, municipal districts, and substations corresponding to each transmission line. Define different voltage-power variation models based on the number of power supply terminals and municipal districts of the transmission line. Combined with the expected power consumption of each municipal district, calculate the expected received power and expected received voltage of each substation on each transmission line. Obtain the actual received power and actual received voltage of each substation on each transmission line to determine the location of the fault on the transmission line. Feedback the location of the fault on the transmission line, build a dual AI operator model, and deploy it in the substation corresponding to each transmission line, and continuously monitor the power and voltage changes on each transmission line.

Images