CN108957253B - An early warning method for insulator pollution flashover caused by irregular high dust - Google Patents

Abstract

The invention discloses an early warning method for insulator pollution flashover caused by irregular high dust. The method and the device can adapt to the production period of an irregular high-dust production enterprise and the rule of dust generation, carry out early warning on the pollution flashover condition of the insulator in the high-dust environment around the irregular high-dust production enterprise, effectively prevent the pollution flashover condition of the peripheral insulator and provide effective guarantee for the safety of a power grid.

Description

An early warning method for insulator pollution flashover caused by irregular high dust

technical field

The invention discloses an early warning method for insulator pollution flashover caused by irregular high dust, and belongs to the technical field of power equipment pollution prediction.

Background technique

In power systems, insulators are important components that mechanically connect conductors with different potentials to each other, and their performance plays a critical role in the safe operation of the entire power transmission system. The catenary insulators are subjected to the real-time test of severe weather such as rain, snow and fog and the industrial dust of surrounding factories and mines. These pollutants will eventually cause a large number of accidents caused by power grid flashovers. Therefore, it is very necessary to detect the pollution flashover of insulators. Clean insulators to prevent accidents.

In order to prevent pollution flashover of insulators, the main methods of current technology include: 1. Formulate an insulator inspection plan and assign special personnel to conduct regular inspection and cleaning. The disadvantage of this method is that the cleaning plan may have a large error with the actual situation. If the planned frequency If the frequency is too high, it will consume a lot of manpower and financial resources. If the frequency is too low, it may lead to untimely cleaning. Second, the X-ray-based measuring device detects the salt density and dust density on the insulator. This method requires the installation of relevant instruments on the line. The cost of insulators is very high, and it is difficult to popularize them on a large scale. Third, the method is to establish a prediction model through intelligent algorithms such as neural networks, support vector machines, decision trees, etc., and to predict the pollution flashover level of insulators by finding the speed rule of pollutant deposition. Early warning, this method can predict the pollution flashover problem of the insulator to a certain extent and notify the workers to clean it in time.

The early warning of the pollution flashover of insulators can timely remind power grid companies to clean and prevent accidents, and it is a good solution to solve the pollution flashover problem of insulators. However, not all dust pollution is regular and linear. For example, grain production enterprises are greatly affected by seasons and time periods, and at the same time, grain production cycle, road transportation efficiency and other aspects are affected. The intelligent model built in 2000 tends to use the average value of dust volume in a larger period of time for prediction, which is easy to miss key periods of rapid dust accumulation, resulting in failure of early warning, so that power companies ignore the most likely period of pollution flashover of insulators and cause accidents. .

Therefore, it is necessary to formulate a special prediction method for the dust situation around the irregular high-dust production enterprises, and incorporate the dust accumulation speed and production dust change law into the prediction link in the form of time periods in the production process. Insulator pollution flashover warning.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the present invention provides an early warning method for insulator pollution flashover caused by irregular high dust. Through this method, a dust accumulation rate operator can be established, and the dust accumulation rate, rainfall and wind speed can be established based on a time period. The relationship between pollution flashover and humidity, and further establish a pollution flashover decision model, which can be used for insulator pollution flashover warning.

An early warning method for insulator pollution flashover caused by irregular high dust according to the present invention includes the following steps:

1. An early warning method for insulator pollution flashover caused by irregular high dust, including the following steps:

S1, collect the daily rainfall, dustfall, wind speed, humidity and pollution level information of the insulators around the irregular high-dust production enterprises, and build the insulator history information table DustHistoryTable; count all the data entries in the DustHistoryTable to obtain the average rainfall AvgF003, dust reduction Volume mean Avg F004, wind speed mean AvgF005, humidity mean AvgF006:

S101, build a DustHistoryTable, the table contains the following fields:

F001: Week, a record in the table corresponds to the week of the year;

F002: Date, indicating that the record corresponds to the day of the week;

F003: rainfall, the rainfall of the corresponding date;

F004: Dust fall amount, the dust fall amount of the corresponding date;

F005: wind speed, the wind speed of the corresponding date;

F006: Humidity, the humidity of the corresponding date;

D001: Insulator pollution level, the level is divided into five levels from 1 to 5, 1 is the lowest level and 5 is the highest level; for an insulator, data is collected every day and a record of DustHistoryTable is constructed and stored in DustHistoryTable in chronological order. ;

S102, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F003 field, and obtain the mean value of rainfall AvgF003 and the standard deviation of rainfall StdF003;

S103, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F004 field, and obtain the mean value of the dustfall amount AvgF004 and the standard deviation of the dustfall amount StdF004;

S104, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F005 field, and obtain the mean value of wind speed AvgF005 and the standard deviation of wind speed StdF005;

S105, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F006 field, and obtain the humidity mean value AvgF006 and the humidity standard deviation StdF006;

S2, construct an irregular dustfall accumulation speed operator SpeedOpterator, specify the date distance parameter DayDistance of the operator, the value range of DayDistance is [7, 14], and the default value is 7;

S201, specify the location MPos recorded in the DustHistoryTable to be processed;

S202, read the MPos record of DustHistoryTable, obtain the value of its F004 field, and store it in the variable TempF004;

S203, set the date distance counter DayDistanceCounter=0, and set the dustfall accumulation index DustIndex=0;

S204, read the record whose position is MPos-DayDistanceCounter in the DustHistoryTable, obtain the value of its F004 field, and store it in the variable CurrentF004;

S205, calculate the value of DustIndex according to the formula, and the corresponding formula is as follows:

S206, DayDistanceCounter=DayDistanceCounter+1;

S207, if DayDistanceCounter>=DayDistance, then go to S208, otherwise go to S204;

S208: Count the contents recorded in the DustHistoryTable from MPos-DayDistance to MPos, calculate the average value of the field F003, and obtain the average rainfall value MPosAvgF003 in the time period;

S209, count the contents recorded in the DustHistoryTable from MPos-DayDistance to MPos, calculate the average value of the field F005, and obtain the average wind speed value MPosAvgF005 of the time period;

S210, count the contents recorded from MPos-DayDistance to MPos in the DustHistoryTable, calculate the average value of the field F006, and obtain the average humidity value MPosAvgF006 of the time period;

S211, calculate the dust accumulation index F003Index of rainfall, and the calculation formula is:

S212, calculate the dust accumulation index F005Index of the wind speed, and the calculation formula is:

S213, calculate the dust accumulation index F006Index of humidity, and the calculation formula is:

S214, output DustIndex, F003Index, F005Index, F006Index;

S3, build an insulator pollution flashover decision table DustDecsionTable:

S301, create a DustDecsionTable table, and the field structure of the DustDecsionTable is as follows;

IF01, input field 1;

IF02, input field 2;

IF03, input field 3;

IF04, input field 4;

IF05, input field 5;

IF06, input field 6;

DF01, output field DF01;

S4, establish the insulator pollution flashover decision table record generator DecsionCreater:

S401, specify the position MPos recorded in the DustHistoryTable to be processed;

S402, for DustHistoryTable, take out each record and put it into CurrentRecord. The field structure of CurrentRecord is the same as that of DustHistoryTable;

S403, create a record DecsionRecord in the DustDecsionTable table, and the field structure in the DecsionRecord is the same as that of the DustDecsionTable;

S404, the values of all fields in the DecisionRecord are set to 0;

S405, set the value of the IF01 field of the DecsionRecord, and the calculation formula is:

S406, set the value of the IF02 field of the DecsionRecord, and the calculation formula is:

S407, calculate the MPos record of DustHistoryTable through SpeedOpterator, and obtain the values of DustIndex, F003Index, F005Index, and F006Index;

S408, the value of the IF03 field of the DecsionRecord=DustIndex, the value of the IF04 field of the DecsionRecord=F003Index, the value of the IF05 field of the DecsionRecord=F005Index, the value of the IF06 field of the DecsionRecord=F006Index;

S409, the temporary storage variable TempDec=0;

S410, if the total number of entries in DustHistoryTable<(MPos+DayDistance), go to S412, otherwise go to S411;

S411, read the records of all entries from MPos to MPos+DayDistance of DustHistoryTable, obtain the maximum value of its D001 field and store it in the temporary storage variable TempDec;

S412, the value of the DF01 field of the DecsionRecord=TempDec;

S413, append DecsionRecord to the end of DustDecsionTable;

S5, for each record in the DustHistoryTable table, call the insulator pollution flashover decision table record generator DecsionCreater to generate the content of the DustDecsionTable table; learn the contents of the DustDecsionTable through the support vector machine algorithm, and obtain the pollution flashover decision model Model:

S501, for each record in the DustHistoryTable, call the insulator pollution flashover decision table record generator DecsionCreater to generate the content of the DustDecsionTable table;

S502, learn the content of DustDecsionTable through the support vector machine algorithm, and obtain the support vector machine model Model;

The fields IF01, IF02, IF03, IF04, IF05, and IF06 of DustDecsionTable are used as the input of SVM, and DF01 is used as the output of SVM, and all the contents of DustDecsionTable are learned through SVM algorithm, and the SVM model Model is obtained;

S6, collect rainfall, dustfall, wind speed and humidity near the insulator every day, insert the content at the end of DustHistoryTable, process the data through DecsionCreater, and use Model to warn of pollution flashover:

S601, a record MyRecord of the DustHistoryTable table is created, and the fields of MyRecord are specified as the following values:

F001 = the current week;

F002 = the current day of the week;

F003 = current rainfall;

F004=Current dust reduction amount;

F005=current wind speed;

F006 = current humidity;

D001 = 0;

S602, append MyRecord to the end of the DustHistoryTable;

S603, use DecsionCreater to process the last data of DustHistoryTable;

S604, for DustDecsionTable, use Model to predict its last record, wherein IF01, IF02, IF03, IF04, IF05, and IF06 of the last record are used as the input of the model;

S605, result=output of Model;

S606, if the result is greater than or equal to 4, go to S607, otherwise go to S608;

S607, pollution flashover may occur, carry out pollution flashover warning; go to S609;

S608, no pollution flashover occurs, and no warning is given;

S609, the prediction process ends.

The beneficial effects of the invention are as follows: an early warning method for insulator pollution flashover caused by irregular high dust is provided, and the dust accumulation rate operator can be established by this method, and the relationship between the dust accumulation rate and the rainfall, wind speed and humidity can be established based on the time period. The relationship between the two, and further establish a pollution flashover decision-making model, which can be used for insulator pollution flashover warning. Through the patent of the present invention, it can adapt to the production cycle of irregular high-dust production enterprises and the rules of dust generation, and give early warning to the pollution flashover of insulators in high-dust environments around irregular high-dust production enterprises, effectively preventing surrounding insulators. The occurrence of pollution flashover can provide an effective guarantee for the security of the power grid.

Description of drawings

Fig. 1 is the F001 field data diagram of the historical information table of this region in Example 1 in 2016;

Fig. 2 is the F002 field data diagram of the historical information table of this region in Example 1 in 2016;

Fig. 3 is the F003 field data diagram of the historical information table of embodiment 1 this area in 2016:

Fig. 4 is the F004 field data diagram of the historical information table of the region in Example 1 in 2016:

Fig. 5 is the F005 field data diagram of the historical information table of this region in Example 1 in 2016;

Fig. 6 is the F006 field data diagram of the historical information table of this region in Example 1 in 2016;

Fig. 7 is the D001 field data diagram of the historical information table of this region in Example 1 in 2016;

Fig. 8 is the IF01 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1;

Fig. 9 is the IF02 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1;

Fig. 10 is the IF03 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1:

Figure 11 is the IF04 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1:

Figure 12 is the IF05 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1;

Figure 13 is the IF06 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1;

Figure 14 is the DF01 field data diagram of the insulator pollution flashover decision table in the region in 2016 calculated by the present invention in Example 1;

Figure 15 shows the situation where pollution flashover may indeed occur after manual inspection in the whole year of 2017;

Fig. 16 is the prediction result diagram of the pollution flashover of the method of the present invention;

FIG. 17 is a diagram of a situation where the method of the present invention predicts that pollution flashover may occur.

Detailed ways

The specific embodiments of the present invention are described by the following examples, but those skilled in the art should understand that this is only an illustration, the protection scope of the present invention is defined by the appended claims, and those skilled in the art should not On the premise of departing from the principle and essence of the present invention, various changes or modifications can be made to these embodiments, and these changes and modifications all fall within the protection scope of the present invention.

Example 1

Take an insulator in a certain area as an example:

1. The first step is to introduce the annual data of the region in 2016 and build a DustHistoryTable. The values of each field in the DustHistoryTable are shown in the following figure: Figure 1 is the weekly data map of the region in 2016; Figure 2 is the The date map of the region's annual data in 2016; Figure 3 is the region's annual data rainfall in 2016; Figure 4 is the region's 2016 annual data dustfall; Figure 5 is the region's 2016 annual data wind speed ; Figure 6 is the annual data humidity in the region in 2016; Figure 7 is the annual insulator pollution level data in the region in 2016; AvgF003=5.7; Avg F004=2.83; AvgF005=2.44; AvgF006=6.08;

2. In the second step, specify DayDistance=7, and construct the SpeedOpterator for the accumulation speed of irregular dustfall;

In the third step, build the DustDecsionTable of the insulator pollution flashover decision table;

In step 4, establish the insulator pollution flashover decision table record generator DecsionCreater;

3. The fifth step is to generate the content of the DustDecsionTable. The content of the DustHistoryTable is as follows: Figure 8 is the weekly data chart of the region in 2016 calculated by the present invention; Figure 9 is the annual data of the region in 2016 calculated by the present invention. Date chart; Figure 10 is the annual data rainfall in the region in 2016 calculated by the present invention; Figure 11 is the annual data dustfall in the region in 2016 calculated by the present invention; Figure 12 is the region in 2016 calculated by the present invention. Figure 13 is the annual data humidity of the region in 2016 calculated by the present invention; Figure 14 is the annual insulator pollution level data in the region in 2016 calculated by the present invention; based on the content of DustHistoryTable, the pollution flashover is obtained Decision model Model;

4. In the sixth step, select the 2017 annual data of the insulator for prediction;

After manual inspection in 2017, it is shown in Figure 15 that pollution flashover may indeed occur (in this chart, 1 means that a pollution flashover warning is required, and 0 means that no pollution flashover warning is required).

The prediction result of the pollution flashover of the present invention is shown in Figure 16; the pollution flashover exists, but the number of times that the method described in the present invention fails to predict is 0; the pollution flashover does not exist, but the method of the present invention predicts that the pollution flashover may occur The situation is shown in Figure 17: only 9 times for the full year data. It can be seen from the above experiments that the method of the present invention can successfully predict the pollution flashover of the insulator; the number of missed reports is 0.

Example 2

The comparison between the calculation method of the present invention and the traditional method:

The data of 20 insulators in 2017 were collected around a grain production enterprise to give early warning to insulator pollution flashover, and compared with traditional neural network and decision tree algorithms based on historical data. The comparison results are as follows:

Conclusion: It can be seen that due to the irregularity of grain production enterprises, the neural network and decision tree methods based on historical data have a greater possibility of underreporting, and the method of the present invention has 0 missed reports, which can be compared Good to prevent pollution flashover from happening.

Pollution flashover warning is carried out on a certain insulator of a grain production enterprise, and three possible situations are determined by manual long-term monitoring. Compare this patented method with traditional neural network and decision tree algorithms based on historical data

Conclusion: It can be seen that the present invention can determine the pollution flashover risk earlier than the traditional method and the manual method.

Claims (1)

1. An early warning method for insulator pollution flashover caused by irregular high dust, comprising the following steps: S1, collect the daily rainfall, dustfall, wind speed, humidity and pollution level information of the insulators around the irregular high-dust production enterprises, and build the insulator history information table DustHistoryTable; count all the data entries in the DustHistoryTable to obtain the average rainfall AvgF003, dust reduction Volume mean Avg F004, wind speed mean AvgF005, humidity mean AvgF006: S101, build a DustHistoryTable, the table contains the following fields: F001: Week, a record in the table corresponds to the week of the year; F002: Date, indicating that the record corresponds to the day of the week; F003: rainfall, the rainfall of the corresponding date; F004: Dust fall amount, the dust fall amount of the corresponding date; F005: wind speed, the wind speed of the corresponding date; F006: Humidity, the humidity of the corresponding date; D001: Insulator pollution level, the level is divided into five levels from 1 to 5, 1 is the lowest level and 5 is the highest level; for an insulator, data is collected every day and a record of DustHistoryTable is constructed and stored in DustHistoryTable in chronological order. ; S102, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F003 field, and obtain the mean value of rainfall AvgF003 and the standard deviation of rainfall StdF003; S103, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F004 field, and obtain the mean value of the dustfall amount AvgF004 and the standard deviation of the dustfall amount StdF004; S104, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F005 field, and obtain the mean value of wind speed AvgF005 and the standard deviation of wind speed StdF005; S105, calculate and count all entry data in the DustHistoryTable, calculate the mean value of the F006 field, and obtain the humidity mean value AvgF006 and the humidity standard deviation StdF006; S2, construct an irregular dustfall accumulation speed operator SpeedOpterator, specify the date distance parameter DayDistance of the operator, the value range of DayDistance is [7, 14], and the default value is 7; S201, specify the location MPos recorded in the DustHistoryTable to be processed; S202, read the MPos record of DustHistoryTable, obtain the value of its F004 field, and store it in the variable TempF004; S203, set the date distance counter DayDistanceCounter=0, and set the dustfall accumulation index DustIndex=0; S204, read the record whose location is MPos-DayDistanceCounter in DustHistoryTable, obtain the value of its F004 field, and store it in the variable CurrentF004; S205, calculate the value of DustIndex according to the formula, and the corresponding formula is as follows:

S206, DayDistanceCounter=DayDistanceCounter+1; S207, if DayDistanceCounter>=DayDistance, then go to S208, otherwise go to S204; S208, count the contents recorded in the DustHistoryTable from MPos-DayDistance to MPos, calculate the average value of the field F003, and obtain the average rainfall value MPosAvgF003 in the time period; S209, count the contents recorded in the DustHistoryTable from MPos-DayDistance to MPos, calculate the average value of the field F005, and obtain the average wind speed value MPosAvgF005 of the time period; S210, count the contents recorded in the DustHistoryTable from MPos-DayDistance to MPos, calculate the average value of the field F006, and obtain the average humidity value MPosAvgF006 of the time period; S211, calculate the dust accumulation index F003Index of rainfall, and the calculation formula is:

S212, calculate the dust accumulation index F005Index of the wind speed, and the calculation formula is:

S213, calculate the dust accumulation index F006Index of humidity, and the calculation formula is:

S214, output DustIndex, F003Index, F005Index, F006Index; S3, build an insulator pollution flashover decision table DustDecsionTable: S301, create a DustDecsionTable table, and the field structure of the DustDecsionTable is as follows; IF01, input field 1; IF02, input field 2; IF03, input field 3; IF04, input field 4; IF05, input field 5; IF06, input field 6; DF01, output field DF01; S4, establish the insulator pollution flashover decision table record generator DecsionCreater: S401, specify the position MPos recorded in the DustHistoryTable to be processed; S402, for DustHistoryTable, take out each record and put it into CurrentRecord. The field structure of CurrentRecord is the same as that of DustHistoryTable; S403, create a record DecsionRecord in the DustDecsionTable table, and the field structure in the DecsionRecord is the same as that of the DustDecsionTable; S404, the values of all fields in the DecisionRecord are set to 0; S405, set the value of the IF01 field of the DecsionRecord, and the calculation formula is:

S406, set the value of the IF02 field of the DecsionRecord, and the calculation formula is:

S407, calculate the MPos record of DustHistoryTable through SpeedOpterator, and obtain the values of DustIndex, F003Index, F005Index, and F006Index; S408, the value of the IF03 field of the DecsionRecord=DustIndex, the value of the IF04 field of the DecsionRecord=F003Index, the value of the IF05 field of the DecsionRecord=F005Index, the value of the IF06 field of the DecsionRecord=F006Index; S409, the temporary storage variable TempDec=0; S410, if the total number of DustHistoryTable record entries < (MPos+DayDistance), go to S412, otherwise go to S411; S411, read the records of all entries from MPos to MPos+DayDistance of DustHistoryTable, obtain the maximum value of its D001 field and store it in the temporary storage variable TempDec; S412, the value of the DF01 field of the DecsionRecord=TempDec; S413, append DecsionRecord to the end of DustDecsionTable; S5, for each record in the DustHistoryTable table, call the insulator pollution flashover decision table record generator DecsionCreater to generate the content of the DustDecsionTable table; use the support vector machine algorithm to learn the content of the DustDecsionTable to obtain the pollution flashover decision model Model: S501, for each record in the DustHistoryTable, call the insulator pollution flashover decision table record generator DecsionCreater to generate the content of the DustDecsionTable table; S502, learn the content of DustDecsionTable through the support vector machine algorithm, and obtain the support vector machine model Model; The fields IF01, IF02, IF03, IF04, IF05, and IF06 of DustDecsionTable are used as the input of SVM, and DF01 is used as the output of SVM. All the contents of DustDecsionTable are learned through SVM algorithm, and the SVM model Model is obtained; S6, collect rainfall, dustfall, wind speed and humidity near the insulator every day, insert the content at the end of DustHistoryTable, process the data through DecsionCreater, and use Model to warn of pollution flashover: S601, a record MyRecord of the DustHistoryTable table is created, and the fields of MyRecord are specified as the following values: F001 = the current week; F002 = the current day of the week; F003 = current rainfall; F004=Current dust reduction amount; F005=current wind speed; F006 = current humidity; D001 = 0; S602, append MyRecord to the end of the DustHistoryTable; S603, use DecsionCreater to process the last data of DustHistoryTable; S604, for DustDecsionTable, use Model to predict its last record, wherein IF01, IF02, IF03, IF04, IF05, and IF06 of the last record are used as the input of the model; S605, result=output of Model; S606, if the result is greater than or equal to 4, go to S607, otherwise go to S608; S607, pollution flashover may occur, carry out pollution flashover warning; go to S609; S608, no pollution flashover occurs, and no warning is given; S609, the prediction process ends.

Images