High-voltage transmission line icing process integrated prediction method

Abstract

The invention relates to a high-voltage transmission line icing process integrated prediction method and belongs to the overhead high-voltage transmission line on-line monitoring and early warning technical field. According to the method, icing process historical data of a high-voltage transmission line on-line monitoring system and corresponding micro meteorological data are utilized to establish an icing process fusion prediction system comprising a data pre processing module, an icing process prediction module and a fusion decision-making module. The method includes the following steps that: dimension reduction processing is performed on the meteorological data based on a principal component analysis method, so that the complexity of modeling can be decreased; a multivariable nonlinear regression model, a support vector machine model, an expert evaluation model and a fuzzy reasoning model are utilized to qualitatively or quantitatively predict the icing process of a power transmission line; and a DS evidence fusion theory is utilized to perform united judgment on four prediction results, and an icing early warning level is confirmed. The high-voltage transmission line icing process integrated prediction method has the following advantages that: a modeling process is simplified; the uncertainty of a single model is eliminated, so that the accuracy of prediction is improved; and effective basis can be provided for relevant departments in decision making of icing disasters.

Description

Technical field: [0001] The invention belongs to the technical field of on-line monitoring and early warning of overhead high-voltage transmission lines, and relates to a comprehensive prediction method for the icing process of high-voltage transmission lines under freezing disasters. Background technique: [0002] The power system is a complex system, and the extensive utilization of electric energy has brought the living standards of human beings to an unprecedented height. However, the occurrence of natural disasters such as earthquakes, typhoons, rainstorms, ice and snow has made the power system face huge challenges, causing large-scale power outages and even system collapses, thus causing huge losses to human production and life. Especially in low-latitude and high-altitude areas such as Yunnan, Guizhou, and Sichuan in my country, freezing disasters are more likely to occur in winter, and icing on high-voltage transmission lines is the most important form of manifestat...

AI Technical Summary

Problems solved by technology

However, the icing process of transmission lines is a multi-variable nonlinear process, which has the characteristics of sudden change, uncertainty, dynamicity and nonlinearity. Its analytical model is not easy to establish, and it is impossible to establish a general model.

Therefore, it is difficult to predict icing through a definite analytical model in practical applications

[0009] 2) The intelligent computing model predicts the icing. Although it does not need to establish a specific functional relationship to truly reflect the relationship between the microclimate and the icing process, it has high requirements for the integrity of the sample.

However, in fact, extreme weather samples are difficult to obtain, so that the training model cannot truly reflect the corresponding relationship between meteorological factors and ice cover

In addition, for the nonlinearity, strong coupling, dynamics and abruptness of the icing process, the traditional methods of artificial intelligence are powerless.

[0010] On the other hand, the operation and maintenance personnel of high-voltage transmission lines who have worked in the field for a long time have accumulated a lot of experience in early warning of freezing disasters, and also provided basic data for the prediction of the icing process of high-voltage transmission lines based on expert systems, but the subjective factors are too strong

[0011] To sum up, it is obviously not enough to use a single prediction model for high-dimensional, nonlinear, and abrupt complex icing processes

Images