CN112114384A - Power transmission line icing occurrence probability forecasting method - Google Patents

Abstract

The invention discloses a method for predicting the occurrence probability of icing on power transmission lines. The method comprises: acquiring three-dimensional multivariate and time-continuous grid point data in atmospheric space, and using the T-mode principal component analysis method to analyze, according to the regional grid point data. The weather classification is based on the circulation field in the middle and lower layers of the troposphere; combined with the daily different types of atmospheric circulation in the area to be forecasted in winter for a certain period of time, a conceptual model for the classification of the weather circulation situation is obtained; it can objectively predict the occurrence probability of icing without Relying on the subjective consciousness of business personnel, improve the prediction accuracy of icing occurrence, facilitate the understanding of the synoptic characteristics of icing occurrence in scientific research and business applications, and provide scientific and technological support for the prediction of icing occurrence on transmission lines.

Description

A Probability Prediction Method of Icing Occurrence in Transmission Lines

technical field

The invention relates to a weather forecast technology, in particular to a method for predicting the occurrence probability of icing on a power transmission line.

Background technique

Icing has always been a serious meteorological disaster for transmission lines in the southern region, directly threatening the operation and maintenance of transmission lines. The icing process is not only controlled by the weather situation (cold wave in the north and water vapor in the south) and altitude, but also affected by various factors such as local meteorological elements.

From January 10, 2008 to February 2, 2008, there was a continuous large-scale low-temperature rain, snow, and freezing disaster weather process in southern my country. Many scholars have analyzed the cause of this process. For example, the research found that the atmospheric circulation was abnormal and found the development of the high-pressure ridge in the Ural Mountains. Strong, the northerly guiding airflow before the ridge transports cold air to the south, and the southwesterly airflow before the trough transports water vapor to the north. The intersection of cold air and warm and humid airflow provides conditions for the occurrence of freezing rain and snow weather. Another example is the abnormal atmospheric circulation, water vapor transport and temperature inversion layer to analyze the causes of the low temperature, snowfall and freezing disasters. Based on the analysis of the circulation classification characteristics of the eight persistent low temperature rain and snow freezing events since 1980, it is believed that the large-scale circulation characteristics of the persistent low temperature rain and snow freezing events in southern my country are mainly single resistance type and double resistance type. There is no systematic and objective method to summarize these subjective weather situation analysis, which are mostly individual cases, and lack representativeness and guidance for the icing forecast of transmission lines.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the above-mentioned deficiencies of the prior art, and to provide a method for predicting the occurrence probability of icing on power lines, which predicts the probability of icing occurrence in an objective method, does not rely on the subjective consciousness of business personnel, and improves the prediction accuracy of icing occurrence.

For achieving the above object, the technical scheme of the present invention is:

A method for predicting the probability of occurrence of icing on a transmission line, comprising:

Obtain three-dimensional multivariate and time-continuous grid data in atmospheric space, and use T-mode principal component analysis to analyze the weather, and classify the weather according to the circulation field in the middle and lower layers of the troposphere in the regional grid data;

Combining with the different types of atmospheric circulations in the area to be forecasted in winter for a certain period of time, a conceptual model of weather circulation situation classification is obtained;

According to the icing estimated by the tensile force, the actual encrypted observation icing and the meteorological element data, a historical database of the icing of the transmission line and the local meteorological conditions that need to be forecasted is established, and the conceptual model of the weather circulation situation classification obtained above is combined. , and statistics the thickness and occurrence probability of icing and the characteristics of meteorological elements under different weather circulation situation classifications;

Calculate the leading, contemporaneous and lagging correlation coefficients of various atmospheric circulation indices and the initial thickness of icing and the maximum icing thickness under the conceptual models of different weather circulation situations, and obtain the pre- and contemporaneous weather patterns and meteorological circulation indices and the power transmission in the area to be forecasted The causal relationship between the line icing events is established, the objective weather type conceptual model of icing occurrence probability and falling area is established, and the regression prediction equation of icing thickness is established;

Using the weather situation field in the next 72 hours given by the three-dimensional multivariate and time-continuous grid data in atmospheric space, and forecasting the occurrence probability and falling area of icing according to the synoptic conceptual model;

According to the weather situation field in the next 72 hours, the circulation indicators and local meteorological elements are calculated, and the regression prediction equation of ice thickness is input to forecast the ice thickness of transmission lines in the next 72 hours.

Further, the synthetic analysis is carried out in combination with the different types of atmospheric circulations in the areas that need to be forecasted in winter for a certain period of time, and the conceptual model of weather circulation situation classification is obtained including:

The objective weather classification was carried out by calculating the 850hPa height field in the daily three-dimensional space grid points from the troposphere to the stratosphere in East Asia during the winter of the past 10 years by using the T-mode principal component analysis method;

Then, the meteorological factors of different types of near-surface temperature, humidity, wind and water vapor transport flux are synthesized, and the intensity of the East Asian trough, the Siberian high pressure index, the intensity index of the western Pacific subtropical high, the area index of the western Pacific subtropical high, and the western Pacific subtropical high are calculated. The atmospheric circulation index at the position of the ridge point is used to obtain a conceptual model of the classification of the synoptic circulation situation.

Further, the three-dimensional multi-variable and time-continuous grid point data in atmospheric space are forecast grid data based on three-dimensional multi-variable and time-continuous atmospheric space output from the numerical model of the European Forecast Center.

Further, the meteorological element data includes:

Location information of icing, including latitude and longitude information, altitude information;

Ice-covering characterization parameters, including ice-covering thickness, ice-covering occurrence time, and ice-covering duration;

Local meteorological elements include temperature, precipitation, wind speed, humidity and air pressure.

Further, the method for classifying weather according to the circulation field in the middle and lower layers of the troposphere according to the regional grid data includes:

The 850hPa geopotential height field is classified by the T-PCA method, and the number of classifications is determined by the evaluation of the explained cluster variance ECV. The calculation formula of ECV is as follows:

where WS is the weather type sum of squares and TS is the total sum of squares:

是元素其到质心的欧氏距离平方：k is the number of weather types, Cj is the classification j in the k class,

is the squared Euclidean distance of the element to the centroid:

天气类型j的估计均值，

为估计总平均值；l is the time step (l=1,2,...,m), Y _il represents each data point,

the estimated mean for weather type j,

is the estimated overall average;

The number of weather classifications is ultimately determined by the increment of ECV, ΔECV:

ΔECV ₌ ECVk-ECVk _-1

When the ΔECV reaches the maximum value, the number k of weather types is determined, indicating that the classification performance is greatly improved and tends to be stable

Compared with the prior art, the present invention has the following beneficial effects:

The transmission line icing occurrence probability prediction method provided by the invention can objectively predict the icing occurrence probability, does not rely on the subjective consciousness of business personnel, improves the prediction accuracy of icing occurrence, and facilitates the understanding of icing occurrence weather in scientific research and business applications It can provide scientific and technological support for the prediction of the occurrence of icing on transmission lines.

Description of drawings

Figure 1 is a flow chart of a method for predicting the probability of occurrence of icing on transmission lines in southern China based on objective weather classification

Fig. 2 shows the change of ΔECV and ECV with the number of weather types in the winter of 2014-2018 by taking the study in Guishan area of China Southern Power Grid as an example.

Figure 3 shows the time series changes of the four objectively determined weather types

Figure 4 shows the circulation configuration of the 850hPa geopotential height field and wind field synthesized by four types of objective weather types

Figure 5 shows the circulation configuration of the 500hPa geopotential height field and wind field synthesized by four types of objective weather types

Figure 6 shows the ice thickness data of different transmission line towers from January 25 to February 1, 2018

Figure 7 shows the representative meteorological elements (temperature) data of different transmission line towers from January 25 to February 1, 2018

Figure 8 is a boxplot of ice thickness under four weather types

Figure 9 shows the icing rate and average maximum icing thickness of the tower under four weather types

Detailed ways

Example:

The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

Referring to FIG. 1 , a method for predicting the probability of occurrence of ice coating on a power transmission line provided in this embodiment specifically includes the following steps:

Step A: Using the latest generation of reanalysis data ERA5 atmospheric space three-dimensional multivariate and time-continuous grid data, through T-mode principal component analysis (T-PCA), according to the regional grid data, the circulation field in the middle and lower troposphere Make weather classifications. This example uses the 850hPa geopotential height field in the daily three-dimensional space grid point of the troposphere to the stratosphere in the East Asian region during the winter (December to January, February of the following year) during the period of 2014-2018, and the number of classifications is explained by The cluster variance (ECV) and ΔECV were evaluated and determined. As shown in Figure 2, based on the evaluation results of the objective classification of the 850hPa geopotential height field, it can be seen that when k=4, the ΔECV reached the maximum value of 0.059, indicating that 2014-2019 When the weather types in winter are divided into four categories, the classification results are relatively stable. Therefore, the weather type identification code for each day within the research period can be given, as shown in Figure 3. These objectively identified weather patterns will be used to match the occurrence of icing on subsequent links and wires.

Step B: synthetic analysis is carried out to the atmospheric circulation of different types every day, as Fig. 4 is the circulation configuration of the 850hPa geopotential height field and the wind field synthesized by four types of objective weather types; Fig. 5 is the 500hPa position of the synthesis of four types of objective weather types. Circulation configuration of potential height field and wind field. Similarly, other elements such as temperature, precipitation and humidity can also be synthesized. From Figure 4 and Figure 5, the Guishan section of China Southern Power Grid can be taken as the target area, which is located behind the 500hPa high-altitude trough under Type1, and is controlled by the anticyclonic circulation at the top of the weak high pressure of 850hPa; Type2 is at the bottom of the 500hPa cold vortex, and its 850hPa There is cold shear in the front, and there is a certain degree of north-south wind shear over Guishan; Type 3 is at the rear of the 500hPa cold vortex, while at the rear of the 850hPa cyclone, the northerly wind is dominant; Anticyclonic circulation. The 500hPa East Asia under Type 4 is an inverted Ω flow pattern, the blocking high pressure is maintained stably, and the continuous and stable configuration of the west high and east low pressure will induce the cold air of Siberia to move southward, and the warm and humid air flow from the ocean will be continuously transported to the southern part of my country. The confluence of air will cause freezing rain and snow weather, and UHV transmission lines in some high-altitude areas are more likely to be covered with ice. It can be seen that the conceptual model of the classification of the synoptic circulation situation can be clearly summarized.

Step C: A historical database of China Southern Power Grid’s transmission line icing and local meteorological conditions that occurred was established using the estimated icing and actual encrypted observation icing and meteorological element data from China Southern Power Grid; Figures 6 and 7 are for 2018 From January 25th to February 1st, the ice thickness data and representative meteorological elements (temperature) data of different transmission line towers are used as examples. Combined with the previous conceptual model of weather circulation situation classification, the thickness and occurrence probability of icing and the characteristics of meteorological elements under different conceptual models of weather circulation situation classification can be counted. Figure 8 shows the icing under four types of weather types. Thickness boxplot, Figure 9 shows the ice coverage rate and average maximum ice thickness of the tower under four weather types. It can be seen that the icing rate of Type4 is the highest, reaching an average of 25.08%, and its average icing thickness is 5.29mm. Under this weather type, the icing rate of Guijia 118 and Guiyi 113, which are almost always located, is as high as 50%. The average ice thickness is 8.49-10.36mm. The average icing rate of Type 1 is the lowest at 7.89%, but its extremely thick icing ratio is extremely high, mainly due to the strong warm and humid airflow and sufficient water vapor. The average icing rates of Type2 and Type3 were 11.93% and 9.17%, and their average icing thicknesses were 4.16mm and 1.73mm. On the whole, the icing situation of Guijia Line 118# and Guiyi Line 113# is more serious than other towers, the icing rate is higher, and the thickness of icing is also larger. May depend on the role of microtopography.

Step D: Based on the analysis of the above steps, establish the connection between the previous and contemporaneous weather patterns and meteorological indicators and the transmission line icing event in southern China, that is, calculate the relationship between each atmospheric circulation index and the initial thickness of the icing and the maximum icing thickness. Leading, contemporaneous and lagging correlation coefficients. The corresponding relationship between the probability of icing occurrence and the synoptic circulation conceptual model of the falling area and the magnitudes of various atmospheric circulation indices in the previous period and the same period is given. The local meteorological elements, altitude, and various atmospheric circulation indices in the previous and contemporaneous periods are used as independent variables, and the ice thickness is used as the dependent variable. The stepwise multiple linear regression method is used to model the dependent and independent variables, and the ice thickness is obtained. Regression forecast models. The meteorological circulation indexes mainly include various circulation intensity indexes as follows: the Siberian high pressure intensity index is the standardized value of the mean sea level pressure of 40°-65°N and 80°-120°E; the East Asian trough intensity index is 25° -45°N, 500hPa height field normalized value at 110°-145°E; since the western Pacific subtropical high is relatively westward during the icing period, the range of the western Pacific subtropical high is selected to be north of 10°N, 90°-160° E, its area index is the number of grid points with a geopotential height of 500hPa>588gpm in the range, the intensity index is the accumulation of the difference between the grid with a geopotential height>588gpm and 587gpm, and the west ridge point is the westernmost position of the 588 isoline Longitude, 850hPa western Pacific subtropical high index is 10°-30°N, 110°-150°E average anomaly value of 850hPa geopotential height.

Step E: Finally, use the numerical model (high-resolution weather forecast and assimilation numerical model) of the European Forecast Center (ECWMF) to output the weather forecast field for the next 72 hours, and the historical summary of the occurrence probability of icing under different atmospheric circulation types and The conceptual model of the falling area was compared to obtain the prediction result of the probability of icing. At the same time, the meteorological circulation indicators, local meteorological elements and altitude in the next 72 hours were calculated and input into the regression forecast equation to obtain the forecasting result of icing thickness.

To sum up, the T-mode principal component analysis (T-PCA) method used in this method is an objective mathematical method based on computer, which can classify the weather according to the circulation field in the middle and lower layers of the troposphere based on the regional grid data, and has better performance. It can objectively predict the probability of icing occurrence without relying on the subjective consciousness of business personnel, improve the prediction accuracy of icing occurrence, facilitate the understanding of the synoptic characteristics of icing occurrence in scientific research and business applications, and provide power transmission Provide scientific and technological support for the prediction of the occurrence of icing on the line.

The above-mentioned embodiments are only to illustrate the technical concept and characteristics of the present invention, and the purpose thereof is to enable those of ordinary skill in the art to understand the content of the present invention and implement them accordingly, and not to limit the protection scope of the present invention. All equivalent changes or modifications made according to the essence of the present invention shall be included within the protection scope of the present invention.

Claims (5)

1. a method for forecasting probability of occurrence of ice coating on power lines, is characterized in that, comprises: Obtain three-dimensional multivariate and time-continuous grid data in atmospheric space, and use T-mode principal component analysis to analyze the weather, and classify the weather according to the circulation field in the middle and lower layers of the troposphere in the regional grid data; Combining with the different types of atmospheric circulations in the area to be forecasted in winter for a certain period of time, a conceptual model of weather circulation situation classification is obtained; According to the icing estimated by the tensile force, the actual encrypted observation icing and the meteorological element data, a historical database of the icing of the transmission line and the local meteorological conditions that need to be forecasted is established, and the conceptual model of the weather circulation situation classification obtained above is combined. , and statistics the thickness and occurrence probability of icing and the characteristics of meteorological elements under different weather circulation situation classifications; Calculate the leading, contemporaneous and lagging correlation coefficients between the atmospheric circulation index and the initial thickness of icing and the maximum icing thickness under the conceptual models of different weather circulation situations, and obtain the pre- and contemporaneous weather patterns and meteorological circulation indexes and the coverage of power lines in the areas to be forecasted. Correlation between ice events, establish an objective conceptual model of the occurrence probability of icing and the weather pattern of the falling area, as well as the regression prediction equation of icing thickness; Using the weather situation field in the next 72 hours given by the three-dimensional multivariate and time-continuous grid data in atmospheric space, and forecasting the occurrence probability and falling area of icing according to the synoptic conceptual model; According to the weather situation field in the next 72 hours, the circulation indicators and local meteorological elements are calculated, and the regression prediction equation of ice thickness is input to forecast the ice thickness of transmission lines in the next 72 hours. 2. The method for predicting the probability of occurrence of icing on power lines as claimed in claim 1, wherein the synthetic analysis is carried out in combination with the different types of atmospheric circulations in areas that need to be forecasted in winter for a certain period of time, and the weather circulation situation classification is obtained. Conceptual models include: The objective weather classification was carried out by calculating the 850hPa height field in the daily three-dimensional space grid points from the troposphere to the stratosphere in East Asia during the winter of the past 10 years by using the T-mode principal component analysis method; Then, the meteorological factors of different types of near-surface temperature, humidity, wind and water vapor transport flux are synthesized, and the intensity of the East Asian trough, the Siberian high pressure index, the intensity index of the western Pacific subtropical high, the area index of the western Pacific subtropical high, and the western Pacific subtropical high are calculated. The atmospheric circulation index at the position of the ridge point is used to obtain a conceptual model of the classification of the synoptic circulation situation. 3. The method for predicting the occurrence probability of icing on transmission lines according to claim 1 or 2, wherein the three-dimensional multivariate and time-continuous grid point data in the atmospheric space is the atmospheric space based on the numerical model output of the European Forecast Center 3D multivariate and time-continuous forecast grid data. 4. The method for predicting the probability of occurrence of ice coating on transmission lines according to claim 1 or 2, wherein the meteorological element data comprises: Location information of icing, including latitude and longitude information, altitude information; Ice-covering characterization parameters, including ice-covering thickness, ice-covering occurrence time, and ice-covering duration; Local meteorological elements include temperature, precipitation, wind speed, humidity and air pressure. 5. The method for predicting the occurrence probability of ice coating on power lines as claimed in claim 1 or 2, wherein the method for weather classification according to the circulation field in the middle and lower layers of the troposphere of the regional grid data comprises: The 850hPa geopotential height field is classified by the T-PCA method, and the number of classifications is determined by the evaluation of the explained cluster variance ECV. The calculation formula of ECV is as follows:

where WS is the weather type sum of squares and TS is the total sum of squares:

k is the number of weather types, Cj is the classification j in the k class,

is the squared Euclidean distance of the element to the centroid:

l is the time step, l = 1, 2, ..., m, Y _il represents each data point,

the estimated mean for weather type j,

is the estimated overall average; The number of weather classifications is ultimately determined by the increment of ECV, ΔECV: ΔECV ₌ ECVk-ECVk _-1 When ΔECV reaches the maximum value, the number k of weather types is determined, indicating that the classification performance is greatly improved and tends to be stable.

Images