Power grid icing prediction method and device

Abstract

The invention provides a power grid icing prediction method and device. The method comprises steps that historical wire icing day number data is acquired, and icing degree data is acquired through calculation according to the wire icing day number data; atmospheric circulation index data is acquired, and characteristic factors are acquired according to relationships between the atmospheric circulation index data and the historical wire icing day number data; the icing degree data and the characteristic factors are inputted to a SVM machine learning model for fitting, and a fitting function model is acquired; atmospheric circulation index data corresponding to to-be-detected data is inputted to the fitting function model, and the power grid icing prediction information is acquired. The method is advantaged in that through integrating the atmospheric circulation index and machine learning methods, a problem that only a few of basic meteorological factors participate into prediction is solved, and effective support is provided for middle-term and long-term icing prediction.

Description

technical field [0001] The invention relates to the field of power grid security, in particular to a method and device for forecasting ice coverage of a power grid. Background technique [0002] The harm of icing to the power grid can be divided into five categories according to its causes: (1) Line overload. In cold, rainy and snowy weather, the ice coating on the conductors continues to grow, resulting in the continuous increase in the quality and volume of the transmission conductors, increasing the sag of the conductors, and reducing the distance to the ground. When the accumulation reaches a certain level, flashover accidents may occur . At the same time, the sag and volume of the wires increase, and under the action of wind force, it may cause the two wires or the wires to collide with the ground, causing accidents such as short-circuit tripping, burns or even burning wires. When the mass of ice coating increases further and exceeds the mechanical strength of wires, ...

AI Technical Summary

Problems solved by technology

(2) Adjacent stalls are not evenly covered with ice or deiced at different times

(3) Insulator string ice flash

(4) The galloping of the transmission wire damages the power equipment

When the wire dances, it will damage the tower, wires, hardware and components, causing frequent trips or even power failures

(5) Icing accidents of substation equipment

This ice disaster caused huge losses to the national economy and people's lives. The direct economic loss of China Southern Power Grid alone reached more than 15 billion yuan. Based on this, how to provide an accurate and effective ice coverage prediction method has become an urgent problem in the industry. question

[0004] Existing icing prediction methods are mainly divided into physical numerical models and statistical analysis methods, both of which only consider basic meteorological elements or factors, rarely add climate scale factors and circulation characteristic fields, and do not take climate Guiding significance of factors and atmospheric circulation on forecasting ice-covered weather

Images