Time Series Prediction and Pattern Recognition of Fault Decomposition Gas for High Voltage Composite Electrical Appliances Based on Grey System OBGM (1,N) Model

Abstract

The GIL/GIS is widely used in the power system. The application of grey system to forecast and analyze the time series of gas decomposition and fault pattern recognition has good engineering value for its operation condition monitoring. Based on the programming of MATLAB software environment, time series prediction model based on grey system OBGM (1,N) is realized. Based on model, grey system time series analysis algorithm is applied to predict the time-varying sequence of the volume fraction of SF6 gas decomposition products and the characteristics of the partial discharge Atlas of the high-voltage combined electrical equipment GIL/GIS in fault state. Parametric time series, grey system is further applied to data mining and analysis of association rules of time series state change, and the time series set is applied to cluster analysis of typical fault types of high-voltage GIL/GIS power equipment. The grey time series prediction and grey relational clustering analysis of small sample test data can be carried out by using the grey system to analyze large data of high voltage GIL/GIS power equipment condition assessment. The multi-dimensional and the grey system information fusion technology proposed in this paper is especially suitable for the application of the small sample and the poor data in the high-voltage GIL/GIS power equipment operation condition detection technology. The time series prediction and the pattern recognition technology of the fault decomposition gas based on the grey system OBGM (1, N) model proposed can provide some technical support for the operation and maintenance of high voltage combinations. The characteristic parameters of partial discharge pattern and grey relational clustering analysis have the good engineering and the theoretical value.