A method for identifying icing hazards based on catenary angle data of transmission lines
By constructing a regression analysis model based on the catenary angle of the transmission line, identifying the ice-covered situation of transmission line, solving the problem of low accuracy and efficiency in the existing technology, and achieving efficient and low-cost ice-covered disaster monitoring to adapt to different Environmental conditions.
Patent Information
- Application Number
- CN202411430212.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-14
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-10-14
AI Technical Summary
The existing transmission line ice recognition methods have low accuracy and efficiency, making them difficult to apply in actual work, especially in remote areas and under changing climate conditions, and the equipment maintenance and data transmission bandwidth requirements are high, resulting in cost and efficiency problems.
Based on the catenary angle data of the transmission line, a catenary angle abnormality identification module is constructed that automatically adapts to the spatial position and overall characteristics. Through the regression analysis model, the transmission line ice-covered situation is identified, which reduces the cost of equipment deployment and maintenance and improves identification efficiency.
It improves the accuracy and efficiency of the identification of disasters in the transmission line ice covering, reduces the requirements for equipment maintenance and data transmission bandwidth, adapts to different climates and construction conditions, is easy to apply in practice, and reduces maintenance costs.
Smart Images

Figure CN118941971B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of transmission line icing identification, and in particular to an icing disaster identification method based on transmission line catenary angle data. Background Art
[0002] In northern regions, transmission lines are often covered with ice due to meteorological disasters such as low temperature, high humidity, and freezing rain. If ice is not dealt with in a timely manner, it will lead to serious accidents such as transmission lines being broken by the gravity of the ice and transmission towers being crushed. Therefore, it is necessary to carry out effective inspections of ice coverage on transmission lines. Rapid and high-quality detection of ice coverage is of great significance to improving the efficiency of transmission line inspections, reducing losses caused by disasters, and improving the level of power grid management.
[0003] At present, the main methods for identifying ice on transmission lines include: (1) manual inspection, where humans arrive at a specific area to inspect the transmission lines. However, since ice is more likely to occur in remote mountainous areas with changeable climates, it is difficult for inspectors to arrive. Therefore, this method is inefficient, costly, and seriously affected by the scope of human activities; (2) drone inspection, where drones are used to inspect transmission lines over a larger area. This method is more efficient than traditional manual inspection and is not affected by the scope of human activities. However, due to the limited cruising range of drones, this method can only cover part of the area and cannot be implemented in remote areas; (3) adding cameras to each section of the transmission line to visually detect ice coverage. This method can cover a larger area and can be implemented in remote areas. However, the camera equipment needs to be maintained to keep the image clear, and real-time image transmission requires a large data transmission bandwidth. The requirements are high, and the maintenance of cameras and broadband will bring huge additional equipment investment and management and maintenance burdens to the transmission line system. If the camera maintenance is not timely or the data transmission broadband does not meet the requirements, it will lead to low efficiency in identifying ice on the transmission line; (4) Detect the angle of the catenary of the power line. If it exceeds the threshold, it means that the line is stretched due to gravity. This method solves the problem of high maintenance equipment costs. However, the catenary angle of the transmission line is affected by many factors such as line material, manufacturer, regional climate, tower quality, construction method, thermal expansion and contraction. Setting a single threshold requires knowing all construction reference factors in the entire area and having extremely high consistency. In actual work, the transmission line in a region is often built gradually over many years and multiple projects. Many factors are unknown and uncontrollable, so this method is difficult to implement in actual work. (5) Early warning based on weather forecasts. In the northern region, some low temperature and high humidity weather is localized in a small area (such as the shady area of some mountains), and weather forecasts are usually large spatial scales and cannot focus on specific small areas, resulting in the inability to accurately identify the ice on the transmission line. Summary of the invention
[0004] The purpose of the present invention is to solve the problems of low recognition accuracy and efficiency and difficulty in application in actual work in the existing transmission line icing disaster recognition methods, and propose an icing disaster recognition method based on transmission line catenary angle data.
[0005] A method for identifying icing disasters based on transmission line catenary angle data, comprising:
[0006] S1. Input the transmission line catenary angle historical data list History, obtain the regional longitude maximum value MaxJ, regional longitude minimum value MinJ, regional latitude maximum value MaxW, regional latitude minimum value MinW, use MaxJ, MinJ, MaxW, MinW to establish the average angle and spatial position decision table HistoryDecision, use the data in HistoryDecision to train the overall catenary average angle and spatial position regression analysis model ValueRegression model, and obtain the trained ValueRegression model;
[0007] S2. Establish a decision support table for reasonable spatial position change interval PostionDecision, and use the data in PostionDecision to train the regression analysis model GapRegression model of the overall catenary average angle and reasonable spatial position change interval to obtain the trained GapRegression model;
[0008] S3. Use the trained ValueRegression model and the trained GapRegression model to construct a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics. The input of the Recognize module is the average value of the catenary angle of the regional transmission line for identification Rjdavg, the regional relative longitude encoding value RPosJ for identification, the regional relative latitude encoding value RPosW for identification, and the real transmission line catenary angle RReal of the point for identification and prediction; the output of the Recognize module is the recognition result ROutput;
[0009] S4. Input the current data list Current of the transmission line catenary angle, use the Recognize module and Current to identify the icing condition of the transmission line, and issue an icing condition warning for the transmission line.
[0010] Furthermore, the input transmission line catenary angle historical data list History in S1 obtains the regional longitude maximum value MaxJ, the regional longitude minimum value MinJ, the regional latitude maximum value MaxW, and the regional latitude minimum value MinW, and uses MaxJ, MinJ, MaxW, and MinW to establish an average angle and spatial position decision table HistoryDecision, and uses the data in HistoryDecision to train the overall catenary average angle and spatial position regression analysis model ValueRegression model to obtain the trained ValueRegression model, specifically:
[0011] S101, input the transmission line catenary angle data history list History;
[0012] Each piece of data in History is the information obtained by a tilt sensor on a regional transmission line;
[0013] Each piece of data in History contains 4 fields:
[0014] (1) Riqi: the date corresponding to the data collection;
[0015] (2) Qingjiao: the inclination angle expressed in radians collected by the inclination sensor;
[0016] (3) WJingdu: the longitude value of the location of the tilt sensor;
[0017] (4) Wweidu: the latitude value of the location of the tilt sensor;
[0018] S102, grouping the data in History based on the Riqi field of History, counting the Qingjiao average value of each group, and obtaining a date statistical average value table HistoryAverage;
[0019] HistoryAverage contains two fields:
[0020] (1) AvgRiqi: the date corresponding to the statistical grouping;
[0021] (2) AvgQingjiao: Statistically obtain the average value of the inclination angle;
[0022] S103, obtaining the maximum longitude value MaxJ of the region, the minimum longitude value MinJ of the region, the maximum latitude value MaxW of the region, and the minimum latitude value MinW of the region:
[0023] MaxJ=the maximum value of the WJingdu field of all data in the History table;
[0024] MinJ=minimum value of WJingdu field of all data in History table;
[0025] MaxW = the maximum value of the Wweidu field of all data in the History table;
[0026] MinW = the minimum value of the Wweidu field of all data in the History table;
[0027] S104, establishing an average angle and spatial position decision table HistoryDecision;
[0028] HistoryDecision contains 4 fields:
[0029] (1) AvgQingjiao: Statistically obtain the average value of the inclination;
[0030] (2) PosJ: relative longitude code value;
[0031] (3) PosW: relative latitude encoding value;
[0032] (4) DecisionQingjiao: inclination decision value;
[0033] S105, set data analysis counter P01Counter=1;
[0034] S106, set the data analysis temporary storage row P01TempRow=the P01Counter row of the History table;
[0035] S107. Create a new row P01NewRow for the HistoryDecision table;
[0036] S108. Get the AvgQingjiao field of P01NewRow:
[0037] Find the row whose AvgRiqi field is equal to the Riqi field value of P01TempRow in the HistoryAverage table, and obtain the value of AvgQingjiao corresponding to the current row;
[0038] S109, obtain the PosJ field of P01NewRow;
[0039] S110, obtain the PosW field of P01NewRow;
[0040] S111, the DecisionQingjiao field of P01NewRow = the Qingjiao field value of P01TempRow;
[0041] S112, set P01Counter=P01Counter+1;
[0042] S113, if P01Counter is less than or equal to the total number of data entries in History, go to S106, otherwise go to S114;
[0043] S114. Establish a regression analysis model ValueRegression of the average angle and spatial position of the overall catenary. The input of the ValueRegression model is the AvgQingjiao, PosJ, and PosW fields of HistoryDecision, and the output is the DecisionQingjiao field.
[0044] S115. Use all the data in HistoryDecision to train the ValueRegression model to obtain a trained ValueRegression model.
[0045] Furthermore, the step of obtaining the PosJ field of P01NewRow in S109 is specifically:
[0046]
[0047] Among them, Round() is used to round the decimal part to the nearest integer.
[0048] Furthermore, the step of obtaining the PosW field of P01NewRow in S110 is specifically:
[0049] Among them, Round() is used to round the decimal part to the nearest integer.
[0050] Furthermore, the establishment of the spatial position reasonable change interval decision support table PostionDecision in S2 uses the data in PostionDecision to train the regression analysis model GapRegression model of the overall catenary average angle and the spatial position reasonable change interval to obtain the trained GapRegression model, specifically:
[0051] S201, group the data of HistoryDecision based on the PosJ and PosW fields of the HistoryDecision table, and count the maximum, minimum and average values of DecisionQingjiao for each group; put the statistical results into the spatial position statistical temporary storage table HistoryPosTStac, TPosJ and TPosW of HistoryPosTStac correspond to the PosJ and PosW fields of the statistical output, and TMaxD, TMinD and TavgD fields correspond to the maximum, minimum and average values of DecisionQingjiao of the statistical output;
[0052] S202, establish a spatial position reasonable change interval decision support table PositionDecision;
[0053] PostionDecision contains 4 fields:
[0054] (1) PDPosJ: spatial position longitude coding of the decision support table for reasonable spatial position change interval;
[0055] (2) PDPosW: spatial position latitude coding of the decision support table for reasonable spatial position change interval;
[0056] (3) PDBt: the lower limit of the interval change of the decision support table of reasonable spatial position change interval;
[0057] (4) PDUp: The upper limit of the interval change of the decision support table for reasonable spatial position change interval;
[0058] S203, establish a counter for reasonable spatial position change data generation PDCounter=1;
[0059] S204, spatial position temporary storage data row TTTempRow=the PDCounterth row of HistoryPosTStac;
[0060] S205, create a new row PDNewRow for the PositionDecision table;
[0061] S206, PDPosJ field of PDNewRow = TTTempRow.TPosJ;
[0062] Among them, TTTempRow.TPosJ is the longitude code of the spatial position stored in TTTempRow;
[0063] S207, PDPosW field of PDNewRow = TTTempRow.TPosW;
[0064] Among them, TTTempRow.TPosW is the spatial position latitude code stored in TTTempRow;
[0065] S208, PDBt field of PDNewRow = TTTempRow.TAvgD-TTTempRow.TMinD;
[0066] Among them, TTTempRow.TAvgD is the average value of DecisionQingjiao stored in TTTempRow, and TTTempRow.TMinD is the minimum value of DecisionQingjiao stored in TTTempRow;
[0067] S209, obtain the PDUp field of PDNewRow;
[0068] S210, set PDCounter=PDCounter+1;
[0069] S211, if PDCounter is less than or equal to the total number of rows of HistoryPosTStac, go to S204, otherwise go to S212;
[0070] S212, establish the regression analysis model GapRegression of the overall catenary average angle and the reasonable change range of spatial position, the input of GapRegression corresponds to PDPosJ and PDPosW of PositionDecision, and the output corresponds to PDBt and PDUp;
[0071] S213. Use the data of PostionDecision to train the GapRegression model to obtain a trained GapRegression model.
[0072] Furthermore, the PDUp field of PDNewRow is obtained in S209, specifically:
[0073] PDUp field of PDNewRow = TTTempRow.TMaxD - TTTempRow.TAvgD
[0074] Among them, TTTempRow.TMaxD is the maximum value of DecisionQingjiao stored in TTTempRow.
[0075] Furthermore, the trained ValueRegression model and the trained GapRegression model are used in S3 to construct a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics. The input of the Recognize module is the mean value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude code value RPosJ for identification, the regional relative latitude code value RPosW for identification, and the real transmission line catenary angle RReal of the point for identification and prediction; the output of the Recognize module is the recognition result ROutput, which is specifically:
[0076] S301, constructing a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics, wherein the input of Recognize is the average value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude code value RPosJ for identification, and the regional relative latitude code value RPosW for identification; and the real transmission line catenary angle RReal of the predicted point is identified;
[0077] S302, connect the input of the trained ValueRegression model to Rjdavg, RPosJ and RPosW;
[0078] S303, connect the input of the trained GapRegression model to RPosJ and RPosW;
[0079] S304, the spatial position angle value RTValue predicted by the model = the output of the trained ValueRegression model;
[0080] S305, the lower limit of the reasonable change interval of the spatial position predicted by the model RTBt and the lower limit of the reasonable change interval of the spatial position predicted by the model RTUp = the output of the trained GapRegression model;
[0081] S306, set the output recognition result variable ROutput=0;
[0082] Among them, ROutput=0 indicates normal state
[0083] S307, identification test first temporary variable RTTemp1=(RTValue-RTBt)-(RTUp-RTBt)*0.10;
[0084] S308, identify the second temporary variable RTTemp2=(RTValue+RTUp);
[0085] S309, identification test third temporary variable RTTemp3=(RTValue+RTUp)+(RTUp-RTBt)*0.10;
[0086] S310, if RReal is less than RTTemp1, set ROutput=-1 and go to S314;
[0087] Among them, ROutput=-1 means that the line or sensor is abnormal;
[0088] S311, if RReal is greater than or equal to RTTemp1 and less than or equal to RTTemp2, set ROutput=0 and go to S314;
[0089] S312, if RReal is greater than RTTemp2 and less than or equal to RTTemp3, set ROutput=1 and go to S314;
[0090] Among them, ROutput=1 means that it is developing into ice-covered state;
[0091] S313, if RReal is greater than RTTemp3, set ROutput=2 and go to S314;
[0092] Among them, ROutput=2 indicates that ice covering occurs;
[0093] S314. Output the value of ROutput as the result of the Recognize module.
[0094] Furthermore, the current data list Current of the catenary angle of the input transmission line in S4 is used to identify the icing condition of the transmission line using the Recognize module and Current, and issue an icing condition warning for the transmission line, specifically:
[0095] S401, input the current data list Current of the transmission line catenary angle;
[0096] The structure of Current is consistent with History;
[0097] S402, current inclination average FuBCQingjiao=average value of Qingjiao field in Current;
[0098] S403, setting the ice disaster identification counter FuBCounter=1;
[0099] S404, the longitude FuBPosJ of the icing disaster identification position is = WJingdu of the FuBCounterth data in the Current table;
[0100] S405, obtaining the icing disaster identification position identification longitude code FuBPosJBM;
[0101] S406, ice disaster identification location identification latitude FuBPosW = Wweidu of the FuBCounterth data in the Current table;
[0102] S407, obtaining the icing disaster identification position identification latitude code FuBPosWBM;
[0103] S408, the current true value FuBValue of ice disaster identification = Qingjiao of the FuBCounterth data in the Current table;
[0104] S409, obtaining icing disaster identification result FuBResult:
[0105] Run the Recognize module, set the module input value to: Rjdavg=FuBCQingjiao, RPosJ=FuBPosJBM, RPosW=FuBPosWBM, RReal=FuBValue, and obtain the icing disaster recognition result FuBResult;
[0106] S410: Output is performed according to the value of FuBValue and an ice-covered wire warning is issued. The output rule is:
[0107] (1) If FuBValue is equal to 0, the output transmission line in the area with longitude FuBPosJ and latitude FuBPosW is normal;
[0108] (2) If FuBValue is equal to 1, the output is that the regional transmission line at longitude FuBPosJ and latitude FuBPosW is developing into an icing disaster, and a disaster warning is issued;
[0109] (3) If FuBValue is equal to 2, it is output that the transmission line in the area with longitude FuBPosJ and latitude FuBPosW has experienced ice disaster, and a disaster warning is issued;
[0110] (4) If FuBValue is equal to -1, the output indicates that the tilt sensor of the regional transmission line at longitude FuBPosJ and latitude FuBPosW is faulty, and an early warning is issued indicating that maintenance is required;
[0111] S411, set FuBCounter=FuBCounter+1;
[0112] S412, if FuBCounter is less than or equal to the total number of data entries in the Current table, go to S404, otherwise go to S413;
[0113] S413, end.
[0114] Furthermore, the acquisition of the ice disaster identification position identification longitude code FuBPosJBM in S405 is specifically:
[0115]
[0116] Among them, Round() is used to round the decimal part to the nearest integer.
[0117] Furthermore, the step of obtaining the ice disaster identification position identification latitude code FuBPosWBM in S407 is specifically:
[0118]
[0119] Among them, Round() is used to round the decimal part to the nearest integer.
[0120] The beneficial effects of the present invention are:
[0121] The present invention proposes a method for identifying icing disasters based on the catenary angle data of transmission lines. The present invention constructs a catenary angle abnormality recognition module Recognize module that automatically adapts to the spatial position and overall characteristics. The Recognize module can be used to adapt to the overall situation and spatial position of the transmission line to find the area where the catenary angle changes abnormally at a specific position, thereby identifying the icing disaster of the transmission line, and improving the accuracy of the identification of the icing disaster of the transmission line. On the one hand, the present invention only judges the icing disaster based on the catenary angle of the transmission line, which is relatively easy to obtain data, and significantly reduces the deployment and maintenance costs of the data collection equipment. At the same time, the present invention is not limited by equipment maintenance and broadband requirements, and improves the efficiency of the identification of icing disasters of the transmission line. On the other hand, the transmission line catenary angle used in the present invention is automatically adaptive, and has nothing to do with the specific transmission line type, construction method, regional climate and other parameters, and is easy to implement and apply in the actual power grid environment. The present invention can improve the inspection efficiency of icing disasters on transmission lines and reduce maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0122] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION
[0123] Specific implementation method 1: Figure 1 As shown, the specific process of the ice disaster identification method based on the transmission line catenary angle data in this embodiment is as follows:
[0124] S1. Input the transmission line catenary angle historical data list History, obtain the regional longitude maximum value MaxJ, regional longitude minimum value MinJ, regional latitude maximum value MaxW, regional latitude minimum value MinW, use MaxJ, MinJ, MaxW, MinW to establish the average angle and spatial position decision table HistoryDecision, use the data in HistoryDecision to train the overall catenary average angle and spatial position regression analysis model ValueRegression model, and obtain the trained ValueRegression model, specifically:
[0125] S101. Input the transmission line catenary angle data history list History. Each piece of History data is information obtained by a tilt sensor on a regional transmission line. Each piece of History data contains 4 fields:
[0126] (1) Riqi: the date corresponding to the data collection;
[0127] (2) Qingjiao: the inclination angle expressed in radians collected by the inclination sensor;
[0128] (3) WJingdu: the longitude value of the location of the tilt sensor;
[0129] (4) Wweidu: the latitude value of the location of the tilt sensor;
[0130] S102. Group the data in History based on the Riqi field of History, count the Qingjiao average value of each group, and obtain the date statistical average value table HistoryAverage. HistoryAverage contains two fields:
[0131] (1) AvgRiqi: the date corresponding to the statistical grouping;
[0132] (2) AvgQingjiao: Statistically obtain the average value of the inclination angle;
[0133] S103, obtain the maximum longitude value MaxJ of the region = the maximum value of the WJingdu field of all data in the History table, the minimum longitude value MinJ of the region = the minimum value of the WJingdu field of all data in the History table, the maximum latitude value MaxW of the region = the maximum value of the Wweidu field of all data in the History table, and the minimum latitude value MinW of the region = the minimum value of the Wweidu field of all data in the History table;
[0134] S104. Create an average angle and spatial position decision table HistoryDecision, which contains 4 fields:
[0135] (1) AvgQingjiao: Statistically obtain the average value of the inclination;
[0136] (2) PosJ: relative longitude code value;
[0137] (3) PosW: relative latitude encoding value;
[0138] (4) DecisionQingjiao: inclination decision value;
[0139] S105, set data analysis counter P01Counter=1;
[0140] S106, set the data analysis temporary storage row P01TempRow=the P01Counter row of the History table;
[0141] S107. Create a new row P01NewRow for the HistoryDecision table;
[0142] S108, AvgQingjiao field of P01NewRow = find the row whose AvgRiqi field is equal to the Riqi field value of P01TempRow in the HistoryAverage table, and obtain the value of AvgQingjiao corresponding to the current row;
[0143] S109. Get the PosJ field of P01NewRow:
[0144]
[0145] Among them, Round() is to round the decimal part to the nearest integer;
[0146] S110. Get the PosW field of P01NewRow:
[0147]
[0148] S111, the DecisionQingjiao field of P01NewRow = the Qingjiao field value of P01TempRow;
[0149] S112, set P01Counter=P01Counter+1;
[0150] S113, if P01Counter is less than or equal to the total number of data entries in History, go to S106, otherwise go to S114;
[0151] S114. Establish the overall catenary average angle and spatial position regression analysis model ValueRegression. The input of the ValueRegression model is the AvgQingjiao, PosJ, and PosW fields of HistoryDecision, and the output is the DecisionQingjiao field.
[0152] The ValueRegression model in the present invention adopts the existing regression analysis model;
[0153] S115. Use all the data in HistoryDecision to train the ValueRegression model to obtain a trained ValueRegression model.
[0154] S2. Establish a decision support table for reasonable spatial position change interval PostionDecision, and use the data in PostionDecision to train the overall catenary average angle and spatial position reasonable change interval regression model GapRegression model to obtain the trained GapRegression model, specifically:
[0155] S201, group the data of HistoryDecision based on the PosJ and PosW fields of the HistoryDecision table, and count the maximum, minimum and average values of DecisionQingjiao for each group; put the statistical results into the spatial position statistical temporary storage table HistoryPosTStac, TPosJ and TPosW of HistoryPosTStac correspond to the PosJ and PosW fields of the statistical output, and TMaxD, TMinD and TavgD fields correspond to the maximum, minimum and average values of DecisionQingjiao of the statistical output;
[0156] S202. Establish a spatial position reasonable change interval decision support table PositionionDecision, which contains 4 fields:
[0157] (1) PDPosJ: spatial position longitude coding of the decision support table for reasonable spatial position change interval;
[0158] (2) PDPosW: spatial position latitude coding of the decision support table for reasonable spatial position change interval;
[0159] (3) PDBt: the lower limit of the interval change of the decision support table of reasonable spatial position change interval;
[0160] (4) PDUp: The upper limit of the interval change of the decision support table for reasonable spatial position change interval;
[0161] S203, establish a counter for reasonable spatial position change data generation PDCounter=1;
[0162] S204, spatial position temporary storage data row TTTempRow=the PDCounterth row of HistoryPosTStac;
[0163] S205, create a new row PDNewRow for the PositionDecision table;
[0164] S206, PDPosJ field of PDNewRow = TTTempRow.TPosJ;
[0165] Among them, TTTempRow.TPosJ is the longitude code of the spatial position stored in TTTempRow;
[0166] S207, PDPosW field of PDNewRow = TTTempRow.TPosW;
[0167] Among them, TTTempRow.TPosW is the spatial position latitude code stored in TTTempRow;
[0168] S208, PDBt field of PDNewRow = TTTempRow.TAvgD-TTTempRow.TMinD;
[0169] Among them, TTTempRow.TAvgD is the average value of DecisionQingjiao stored in TTTempRow, and TTTempRow.TMinD is the minimum value of DecisionQingjiao stored in TTTempRow;
[0170] S209, PDUp field of PDNewRow = TTTempRow.TMaxD-TTTempRow.TAvgD;
[0171] Among them, TTTempRow.TMaxD is the maximum value of DecisionQingjiao stored in TTTempRow;
[0172] S210, set PDCounter=PDCounter+1;
[0173] S211, if PDCounter is less than or equal to the number of rows of HistoryPosTStac, go to S204, otherwise go to S212;
[0174] S212, establish the regression analysis model GapRegression of the overall catenary average angle and the reasonable change range of spatial position, the input of GapRegression corresponds to PDPosJ and PDPosW of PositionDecision, and the output corresponds to PDBt and PDUp;
[0175] S213. Use the data from PostionDecision to train the GapRegression model.
[0176] The GapRegression model in the present invention adopts the existing regression analysis model.
[0177] S3. Use the trained ValueRegression model and the trained GapRegression model to construct a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics. The input of the Recognize module is the mean value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude encoding value RPosJ for identification, the regional relative latitude encoding value RPosW for identification, and the real transmission line catenary angle RReal of the point for identification and prediction; the output of the Recognize module is the recognition result ROutput, which is:
[0178] S301, constructing a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics, wherein the input of Recognize is the average value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude code value RPosJ for identification, and the regional relative latitude code value RPosW for identification; and the real transmission line catenary angle RReal of the predicted point is identified;
[0179] S302, connect the input of the trained ValueRegression model to Rjdavg, RPosJ and RPosW;
[0180] S303, connect the input of the trained GapRegression model to RPosJ and RPosW;
[0181] S304, the spatial position angle value RTValue predicted by the model = the output of the trained ValueRegression model;
[0182] S305, the lower limit of the reasonable change interval of the spatial position predicted by the model RTBt and the lower limit of the reasonable change interval of the spatial position predicted by the model RTUp = the output of the trained GapRegression model;
[0183] S306, set the output recognition result variable ROutput=0;
[0184] The identification result variable ROutpu has three states: 0 indicates a normal state; 1 indicates that the transmission line catenary angle is large and is developing into an ice-covered state; 2 indicates that the transmission line catenary angle is too large and is obviously ice-covered; -1 indicates that the transmission line catenary angle is too small and there is an abnormality in the line or sensor;
[0185] S307, identification test first temporary variable RTTemp1=(RTValue-RTBt)-(RTUp-RTBt)*0.10;
[0186] S308, identify the second temporary variable RTTemp2=(RTValue+RTUp);
[0187] S309, identification test third temporary variable RTTemp3=(RTValue+RTUp)+(RTUp-RTBt)*0.10;
[0188] S310, if RReal is less than RTTemp1, set ROutput=-1 and go to S314;
[0189] S311, if RReal is greater than or equal to RTTemp1 and less than or equal to RTTemp2, set ROutput=0 and go to S314;
[0190] S312, if RReal is greater than RTTemp2 and less than or equal to RTTemp3, set ROutput=1 and go to S314;
[0191] S313, if RReal is greater than RTTemp3, set ROutput=2 and go to S314;
[0192] S314. Output the value of ROutput as the result of the Recognize module.
[0193] S4. Input the current data list of the transmission line catenary angle Current. The structure of Current is consistent with History. Use the Recognize module and Current to identify the icing of the transmission line and issue an early warning. Specifically:
[0194] S401, input the current data list Current of the transmission line catenary angle, the structure of Current is consistent with History;
[0195] S402, current inclination angle average FuBCQingjiao=average value of Qingjiao field in Current;
[0196] S403, setting the ice disaster identification counter FuBCounter=1;
[0197] S404, the longitude FuBPosJ of the icing disaster identification position is = WJingdu of the FuBCounterth data in the Current table;
[0198] S405, ice disaster identification location identification longitude coding FuBPosJBM:
[0199]
[0200] Among them, Round() is to round the decimal part to the nearest integer;
[0201] S406, ice disaster identification location identification latitude FuBPosW = Wweidu of the FuBCounterth data in the Current table;
[0202] S407, obtain the icing disaster identification position identification latitude code FuBPosWBM:
[0203]
[0204] Among them, Round is to round the decimal part to the nearest integer;
[0205] S408, the current true value FuBValue of ice disaster identification = Qingjiao of the FuBCounterth data in the Current table;
[0206] S409, ice disaster identification result FuBResult = run the Recognize module, and the module input value is set to: Rjdavg = FuBCQingjiao, RPosJ = FuBPosJBM, RPosW = FuBPosWBM, RReal = FuBValue;
[0207] S410, output according to the value of FuBValue, the output rule is:
[0208] (1) If FuBValue is equal to 0, the output transmission line in the area with longitude FuBPosJ and latitude FuBPosW is normal;
[0209] (2) If FuBValue is equal to 1, the output is that the regional transmission line at longitude FuBPosJ and latitude FuBPosW is developing into an icing disaster, and a disaster warning is issued;
[0210] (3) If FuBValue is equal to 2, then the output is that the transmission line in the area with longitude FuBPosJ and latitude FuBPosW has experienced icing disaster, and a disaster warning is issued;
[0211] (4) If FuBValue is equal to -1, the output indicates that the tilt sensor of the regional transmission line at longitude FuBPosJ and latitude FuBPosW is faulty, and an early warning is issued indicating that maintenance is required;
[0212] S411, set FuBCounter=FuBCounter+1;
[0213] S412, if FuBCounter is less than or equal to the total number of data entries in the Current table, go to S404, otherwise go to S413;
[0214] S413, end.
Claims
1. A method for identifying icing disasters based on transmission line catenary angle data, characterized in that The specific process of the method is: S1. Input the transmission line catenary angle historical data list History, obtain the regional longitude maximum value MaxJ, regional longitude minimum value MinJ, regional latitude maximum value MaxW, regional latitude minimum value MinW, use MaxJ, MinJ, MaxW, MinW to establish the average angle and spatial position decision table HistoryDecision, use the data in HistoryDecision to train the overall catenary average angle and spatial position regression analysis model ValueRegression model, and obtain the trained ValueRegression model; S2. Establish a decision support table for reasonable spatial position change interval PostionDecision, and use the data in PostionDecision to train the regression analysis model GapRegression model of the overall catenary average angle and reasonable spatial position change interval to obtain the trained GapRegression model; S3. Use the trained ValueRegression model and the trained GapRegression model to construct a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics. The input of the Recognize module is the average value of the catenary angle of the regional transmission line for identification Rjdavg, the regional relative longitude encoding value RPosJ for identification, the regional relative latitude encoding value RPosW for identification, and the real transmission line catenary angle RReal of the point for identification and prediction; the output of the Recognize module is the recognition result ROutput; S4. Input the current data list Current of the transmission line catenary angle, use the Recognize module and Current to identify the icing condition of the transmission line, and issue an icing condition warning for the transmission line.
2. The method for identifying icing disasters based on transmission line catenary angle data according to claim 1, characterized in that: The input transmission line catenary angle historical data list History in S1 obtains the regional longitude maximum value MaxJ, the regional longitude minimum value MinJ, the regional latitude maximum value MaxW, and the regional latitude minimum value MinW, and uses MaxJ, MinJ, MaxW, and MinW to establish an average angle and spatial position decision table HistoryDecision, and uses the data in HistoryDecision to train the overall catenary average angle and spatial position regression analysis model ValueRegression model to obtain the trained ValueRegression model, specifically: S101, input the transmission line catenary angle data history list History; Each piece of data in History is information obtained by a tilt sensor on a regional transmission line; Each piece of data in History contains 4 fields: (1) Riqi: the date corresponding to the data collection; (2) Qingjiao: the inclination angle expressed in radians collected by the inclination sensor; (3) WJingdu: the longitude value of the location of the tilt sensor; (4) Wweidu: the latitude value of the location of the tilt sensor; S102, grouping the data in History based on the Riqi field of History, counting the Qingjiao average value of each group, and obtaining a date statistical average value table HistoryAverage; HistoryAverage contains two fields: (1) AvgRiqi: the date corresponding to the statistical grouping; (2) AvgQingjiao: Statistically obtain the average value of the inclination angle; S103, obtaining the maximum longitude value MaxJ of the region, the minimum longitude value MinJ of the region, the maximum latitude value MaxW of the region, and the minimum latitude value MinW of the region: MaxJ=the maximum value of the WJingdu field of all data in the History table; MinJ=minimum value of WJingdu field of all data in History table; MaxW = the maximum value of the Wweidu field of all data in the History table; MinW = the minimum value of the Wweidu field of all data in the History table; S104, establishing an average angle and spatial position decision table HistoryDecision; HistoryDecision contains 4 fields: (1) AvgQingjiao: Statistically obtain the average value of the inclination; (2) PosJ: relative longitude code value; (3) PosW: relative latitude encoding value; (4) DecisionQingjiao: inclination decision value; S105, set data analysis counter P01Counter=1; S106, set the data analysis temporary storage row P01TempRow=the P01Counter row of the History table; S107. Create a new row P01NewRow for the HistoryDecision table; S108. Get the AvgQingjiao field of P01NewRow: Find the row whose AvgRiqi field is equal to the Riqi field value of P01TempRow in the HistoryAverage table, and obtain the value of AvgQingjiao corresponding to the current row; S109, obtain the PosJ field of P01NewRow; S110, obtain the PosW field of P01NewRow; S111, the DecisionQingjiao field of P01NewRow = the Qingjiao field value of P01TempRow; S112, set P01Counter=P01Counter+1; S113, if P01Counter is less than or equal to the total number of data entries in History, go to S106, otherwise go to S114; S114. Establish a regression analysis model ValueRegression of the average angle and spatial position of the overall catenary. The input of the ValueRegression model is the AvgQingjiao, PosJ, and PosW fields of HistoryDecision, and the output is the DecisionQingjiao field. S115. Use all the data in HistoryDecision to train the ValueRegression model to obtain a trained ValueRegression model.
3. The method for identifying icing disasters based on transmission line catenary angle data according to claim 2, characterized in that: The step of obtaining the PosJ field of P01NewRow in S109 is specifically: Among them, Round() is an operation to round off the decimal part.
4. The method for identifying icing disasters based on transmission line catenary angle data according to claim 3, characterized in that: The step of obtaining the PosW field of P01NewRow in S110 is specifically: Among them, Round() is an operation to round off the decimal part.
5. The method for identifying icing disasters based on transmission line catenary angle data according to claim 4, characterized in that: The establishment of the spatial position reasonable change interval decision support table PostionDecision in S2 uses the data in PostionDecision to train the regression analysis model GapRegression model of the overall catenary average angle and the spatial position reasonable change interval to obtain the trained GapRegression model, specifically: S201, group the data of HistoryDecision based on the PosJ and PosW fields of the HistoryDecision table, and count the maximum, minimum and average values of DecisionQingjiao for each group; put the statistical results into the spatial position statistical temporary storage table HistoryPosTStac, TPosJ and TPosW of HistoryPosTStac correspond to the PosJ and PosW fields of the statistical output, and TMaxD, TMinD and TavgD fields correspond to the maximum, minimum and average values of DecisionQingjiao of the statistical output; S202, establish a spatial position reasonable change interval decision support table PositionDecision; PostionDecision contains 4 fields: (1) PDPosJ: spatial position longitude coding of the decision support table for reasonable spatial position change interval; (2) PDPosW: spatial position latitude coding of the decision support table for reasonable spatial position change interval; (3) PDBt: the lower limit of the interval change of the decision support table of reasonable spatial position change interval; (4) PDUp: The upper limit of the interval change of the decision support table for reasonable spatial position change interval; S203, establish a counter for reasonable spatial position change data generation PDCounter=1; S204, spatial position temporary storage data row TTTempRow=the PDCounterth row of HistoryPosTStac; S205, create a new row PDNewRow for the PositionDecision table; S206, PDPosJ field of PDNewRow = TTTempRow.TPosJ; Among them, TTTempRow.TPosJ is the longitude code of the spatial position stored in TTTempRow; S207, PDPosW field of PDNewRow = TTTempRow.TPosW; Among them, TTTempRow.TPosW is the spatial position latitude code stored in TTTempRow; S208, PDBt field of PDNewRow = TTTempRow.TAvgD-TTTempRow.TMinD; Among them, TTTempRow.TAvgD is the average value of DecisionQingjiao stored in TTTempRow, and TTTempRow.TMinD is the minimum value of DecisionQingjiao stored in TTTempRow; S209, obtain the PDUp field of PDNewRow; S210, set PDCounter=PDCounter+1; S211, if PDCounter is less than or equal to the total number of rows of HistoryPosTStac, go to S204, otherwise go to S212; S212, establish the regression analysis model GapRegression of the overall catenary average angle and the reasonable change range of spatial position, the input of GapRegression corresponds to PDPosJ and PDPosW of PositionDecision, and the output corresponds to PDBt and PDUp; S213. Use the data of PostionDecision to train the GapRegression model to obtain a trained GapRegression model.
6. The method for identifying icing disasters based on transmission line catenary angle data according to claim 5, characterized in that: The PDUp field of PDNewRow is obtained in S209, specifically: PDUp field of PDNewRow = TTTempRow.TMaxD - TTTempRow.TAvgD Among them, TTTempRow.TMaxD is the maximum value of DecisionQingjiao stored in TTTempRow.
7. The method for identifying icing disasters based on transmission line catenary angle data according to claim 6, characterized in that: In S3, the trained ValueRegression model and the trained GapRegression model are used to construct the catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics. The input of the Recognize module is the mean value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude code value RPosJ for identification, the regional relative latitude code value RPosW for identification, and the real transmission line catenary angle RReal of the point for identification and prediction; the output of the Recognize module is the recognition result ROutput, which is specifically: S301, constructing a catenary angle anomaly recognition module Recognize that automatically adapts to the spatial position and overall characteristics, wherein the input of Recognize is the average value Rjdavg of the catenary angle of the regional transmission line for identification, the regional relative longitude code value RPosJ for identification, and the regional relative latitude code value RPosW for identification; and the real transmission line catenary angle RReal for identifying the predicted point; S302, connect the input of the trained ValueRegression model to Rjdavg, RPosJ and RPosW; S303, connect the input of the trained GapRegression model to RPosJ and RPosW; S304, the spatial position angle value RTValue predicted by the model = the output of the trained ValueRegression model; S305, the lower limit of the reasonable change interval of the spatial position predicted by the model RTBt and the lower limit of the reasonable change interval of the spatial position predicted by the model RTUp = the output of the trained GapRegression model; S306, set the output recognition result variable ROutput=0; Among them, ROutput=0 indicates normal state; S307, identification test first temporary variable RTTemp1=(RTValue-RTBt)-(RTUp-RTBt)*0.10; S308, identify the second temporary variable RTTemp2=(RTValue+RTUp); S309, identification test third temporary variable RTTemp3=(RTValue+RTUp)+(RTUp-RTBt)*0.10; S310, if RReal is less than RTTemp1, set ROutput=-1 and go to S314; Among them, ROutput=-1 means that the line or sensor is abnormal; S311, if RReal is greater than or equal to RTTemp1 and less than or equal to RTTemp2, set ROutput=0 and go to S314; S312, if RReal is greater than RTTemp2 and less than or equal to RTTemp3, set ROutput=1 and go to S314; Among them, ROutput=1 means that it is developing into ice-covered state; S313, if RReal is greater than RTTemp3, set ROutput=2 and go to S314; Among them, ROutput=2 indicates that ice covering occurs; S314. Output the value of ROutput as the result of the Recognize module.
8. The method for identifying icing disasters based on transmission line catenary angle data according to claim 7, characterized in that: The input transmission line catenary angle current data list Current in S4 uses the Recognize module and Current to identify the transmission line icing situation and issue an icing warning for the power line, specifically: S401, input the current data list Current of the transmission line catenary angle; The structure of Current is consistent with History; S402, current inclination average FuBCQingjiao=average value of Qingjiao field in Current; S403, setting the ice disaster identification counter FuBCounter=1; S404, the longitude FuBPosJ of the icing disaster identification position is = WJingdu of the FuBCounterth data in the Current table; S405, obtaining the icing disaster identification position identification longitude code FuBPosJBM; S406, ice disaster identification location identification latitude FuBPosW = Wweidu of the FuBCounterth data in the Current table; S407, obtaining the icing disaster identification position identification latitude code FuBPosWBM; S408, the current true value FuBValue of ice disaster identification = Qingjiao of the FuBCounterth data in the Current table; S409, obtaining icing disaster identification result FuBResult: Run the Recognize module, set the module input value to: Rjdavg=FuBCQingjiao, RPosJ=FuBPosJBM, RPosW=FuBPosWBM, RReal=FuBValue, and obtain the icing disaster recognition result FuBResult; S410: Output is performed according to the value of FuBValue and an ice-covered wire warning is issued. The output rule is: (1) If FuBValue is equal to 0, the output transmission line in the area with longitude FuBPosJ and latitude FuBPosW is normal; (2) If FuBValue is equal to 1, the output is that the regional transmission line at longitude FuBPosJ and latitude FuBPosW is developing into an icing disaster, and a disaster warning is issued; (3) If FuBValue is equal to 2, it is output that the transmission line in the area with longitude FuBPosJ and latitude FuBPosW has experienced ice disaster, and a disaster warning is issued; (4) If FuBValue is equal to -1, the output indicates that the tilt sensor of the regional transmission line at longitude FuBPosJ and latitude FuBPosW is faulty, and an early warning is issued indicating that maintenance is required; S411, set FuBCounter=FuBCounter+1; S412, if FuBCounter is less than or equal to the total number of data entries in the Current table, go to S404, otherwise go to S413; S413, end.
9. The method for identifying icing disasters based on transmission line catenary angle data according to claim 8, characterized in that: The step of obtaining the ice disaster identification position identification longitude code FuBPosJBM in S405 is specifically: Among them, Round() is used to round the decimal part to the nearest integer.
10. The method for identifying icing disasters based on transmission line catenary angle data according to claim 9, characterized in that: The step S407 of obtaining the ice disaster identification position identification latitude code FuBPosWBM is specifically: Among them, Round() is used to round the decimal part to the nearest integer.
