Computer-Aided System for Wind Turbine Data Analysis

Abstract

Context: The current work on wind turbine failure detection focuses on researching suitable signal processing algorithms and developing efficient diagnosis algorithms. The laboratory research would involve large and complex data, and it can be a daunting task.
Aims: To develop a Computer-Aided system for assisting experts to conduct an efficient laboratory research on wind turbine data analysis. System is expected to provide data visualization, data manipulation, massive data processing and wind turbine failure detection.
Method: 50G off-line SCADA data and 4 confident diagnosis algorithms were used in this project. Apart from the instructions from supervisor, this project also gained help from two experts from Engineering Department. Java and Microsoft SQL database were used to develop the system.
Results: Data visualization provided 6 different charting solutions and together with robust user interactions. 4 failure diagnosis solutions and data manipulations were provided in the system. In addition, dedicated database server and Matlab API with Java RMI were used to resolve the massive data processing problem.
Conclusions: Almost all of the deliverables were completed. Friendly GUI and useful functionalities make user feel more comfortable. The final product does enable experts to conduct an efficient laboratory research. The end of this project also gave some potential extensions of the system.