Correlation Analysis Of Power Quality Time Series And Governance Strategy Based On Data Driven

The increase of node and capacity and the random fluctuations in output power of distributed generation and the access of power electronic equipment has led to a complex grid structure and outstanding problem of power quality in power system.Based on the vigorous development of China's smart grid,the gradual maturity of network communication technology,in order to quickly and efficiently find and analyze valuable information and knowledge from a large amount of power quality time series data,further explore more reasonable power quality governance plan to improve the insufficiency of traditional local governance methods and achieve global governance of power quality pollution,from the perspective of time series correlation analysis and whole network governance,this article proposes a regional governance strategy of power quality based on data-driven.First,data processing is performed on the original time series of power quality recorded by the monitoring network.In the vertical and horizontal dimensions,the important feature points in the original sequence are extracted using the vertical distance and time span,and the sequence is represented in a piecewise linear representation.For the situation that the original sequence under the same power quality index of each node is different after pattern representation,a interpolation optimization method is proposed,each segment sequence is interpolated and optimized,then a sequence array of equal time dimension is constructed.Secondly,the amount of pollution features carried by the segment sequence at each node is not the same,applying data-driven modeling idea and grey relational theory is used to analyze the correlation of node data in sequence arrays,the relational degree between nodes is calculated to determine the coupling extent of the combined effects of pollution sources.According to the node power quality relational degree,the network is divided into multiple control regions,the dominant point of each region is identified,and power quality governance of the whole network is achieved by regional control.Finally,under the simulation environment of Matlab/Simulink,an IEEE 14-node complex disturbance model under the joint action of multiple pollution sources was constructed.Through the extensive processing and analysis of the recorded data,the accuracy of dividing the control area and the rationality of improving power quality of the entire network in each region are verified.The simulition result shows that the use of node data correlation to disperse control region of the network,by adjusting the power quality of the dominant node within the region,the quality of power supply of the whole region can be improved,and multi-region collaborative control can make the whole network power quality pollution is effectively controlled.