Data noise processing method and processing system of oil temperature prediction system for phase modifier

Abstract

The invention discloses a data noise processing method of an oil temperature prediction system for a phase modifier. The method comprises the following steps: collecting oil temperature historical data of the phase modifier; performing overall discrimination on the oil temperature historical data of the phase modifier based on a support vector machine; establishing a linear interpolation method model based on weighted optimization; setting an oil temperature historical data burr point discrimination threshold of the phase modifier based on a linear interpolation principle; accurately judging oil temperature historical data burr point positions and quantity of the phase modifier. and smoothing the historical oil temperature data of the phase modifier based on correlation weighting. The oiltemperature data burr point discrimination accuracy and data preprocessing precision are effectively improved, the prediction precision of the oil temperature prediction system for the phase modifieris ensured, and a powerful support is provided for safe operation of the phase modifier.

Description

technical field [0001] The invention relates to a data noise processing method of a condenser oil temperature prediction system, and belongs to the technical field of condenser oil temperature prediction. Background technique [0002] With the rapid development of long-distance and large-capacity DC transmission technology, large-scale construction of UHV DC converter stations has been carried out nationwide. According to statistics, by the end of 2020, China will build a "four vertical and six horizontal" UHV main grid and 27 UHV DC lines, fully realizing a strong smart grid with UHV as the backbone grid and coordinated development of power grids at all levels. However, DC transmission requires a large amount of reactive power support. In order to meet the demand for reactive power, the use of synchronous condensers to realize reactive power compensation is proposed for the first time in China. The synchronous condenser has fast dynamic reactive power adjustment and short-...

AI Technical Summary

Problems solved by technology

However, in actual work, errors in historical oil temperature data (called data glitches) may occur due to factors such as personnel errors and system failures. In order to remove data glitches in time and improve oil temperature prediction accuracy, it is generally necessary to temperature data for preprocessing

At present, there are many data noise processing methods, mainly including threshold discrimination method, longitudinal comparison method, linear interpolation method, etc., but these methods have problems such as fixed discrimination threshold and lack of smoothing. In order to solve the above problems, it is urgent to study a new The data noise processing method of

Images