Research On The Method Of Power Quality Change Trend Forecasting And Early-warning And Its Application

With the extensive application of interference load of nonlinear, impact, imbalance and so on, as well as sensitivity load in grid load, power quality problems are increasingly serious, and it’s becoming more and more important to do the forecasting and early-warning for power quality. At present, power quality online monitoring technologies are maturing, but deep analysis and functional development of monitoring data still inadequate. Facing massive power quality monitoring data, scientific forecasting is the basis and guarantee of the correct decision-making, the early-warning mechanism based on the scientific forecasting can help identify potential safety hazard in time, and carry on the targeted and effective governance of the power quality problems. So a research on the method of power quality change trend forecasting and early-warning was carried on in this paper.First of all, an optimization combination forecasting method of power quality based on grey forecasting model, linear regression model and random time series model was designed, it’s convenient to predict the state of power quality indexes in the future period of time rapidly and accurately.Secondly, reference power quality national standard, the warning grades and early-warning thresholds dividing system of five steady-state indicators of power quality were established on the basis of the above research work, and the traditional graded early-warning mechanism of power quality based on the historical data and forecast data was formulated.Then, an early-warning method of power quality considering the characteristics of fuzzy clustering was proposed for the shortcoming that the existing early-warning methods of power quality to set warning thresholds are rigid.This early-warning method forms cooperation with power quality national standards, and it’s the depth of excavation and supplementary analysis of the power quality warning conditions after the judgment for “super-national standard” or “do not exceed the national standard”. In addition, the method of setting power quality graded warning thresholds based on fuzzy clustering was also designed, forming a contrast with the method of traditional power quality graded warning in this paper.Finally, an example analysis of forecasting and early-warning was done. The results of example analysis verified the practicability and effectiveness of the method of power quality change trend forecasting and early-warning in this paper.