Research On Visual Analysis For Wind Turbines Monitoring Data

In recent years, the wind powers is developing rapidly and the measured sites and communication devices of wind farms are increasing quickly. The monitoring data of wind turbines has grown exponentially. As a result,the problem of data disaster of the huge monitoring data has happened to the wind farm operation and management personnel. The study of visualization methods is for the real-time, historical monitoring data of wind turbines. It can make the display of the monitoring data more clear and intuitive. It is significant for wind farm operation and management personnel to find the potential pitfalls information of wind turbines totally and it is also important to improve the operational reliability of wind turbines.An intuitive and effective visualization method is proposed based on the timing, multidimensional and fast monitoring data of wind turbines. However, the visualization method of the monitoring data has the problem of less intuition and poor interaction. So the scatter plot matrix visualization method, parallel graph visualization method, and the cooperative multi-view collaborative visualization are brought.In order to obtain a good visual effects, the monitoring data of wind turbines is preprocessed before it is visualized by the visualization method. The Random Forests and PCA are proposed to process the monitoring data. It can obtain data distribution structure based on full integration sample data category information. The purpose to assist the visualization method is realized.For enhancing the intuition of visualization, the Scatter plot matrix visualization method and parallel graph visualization method are brought. Firstly, the roughly correlation is determined by the Scatter plot matrix. Secondly, the complete picture and related properties of the entire monitoring data set is displayed by Parallel graph. In this way, the intuition of visualization is improved by the display.In order to enhance the interaction of visualization, the cooperative multi-view collaborative visualization method is proposed. It is consisted of the scatter plot matrix, scatter plot and the parallel graph. The analysis of cooperative multi-view collaborative visualization is based on the parallel interactive technology and the interactive way of Overview and Detail. The defects of a single visualization method are overcame by the cooperative multi-view collaborative visualization. And the advantages of these visualization methods are reserved. All these can improve the users’ engagement and the interactivity.On the basis of the above visualization methods, considering the features of the wind turbines monitoring data, the results of visualization application in the wind turbines monitoring data are shown in the thesis. The property correlation, the data set characteristics and the distribution of the monitoring data are displayed and analyzed by the way of graph and image.