Pradeep Misra (Committee Member), Kuldip Rattan (Committee Member), Xiaodong Zhang (Committee Chair)
Master of Science in Engineering (MSEgr)
Because wind is a clean and renewable energy, it is in high demand as an alternative resource of fossil fuels for our daily power supply. Thus, wind turbine technology is wildly applied to convert the wind energy to electrical power. With the size of the wind turbines increased, the maintenance cost of the wind turbines also goes up. Fault diagnosis can significantly help wind turbines to: reduce the maintenance cost, the machinery breakdown, spare parts inventories, total machine downtime, and overtime expenses, and increase machine life, overall productivity, and profit. The technical challenges for fault diagnosis of the wind turbine are that the wind itself is hardly measureable, and the aerodynamics of the turbine is nonlinear. In this research we will present a robust observer-based fault diagnosis method for wind turbines.
A benchmarked model of wind turbine for fault diagnosis is considered in this thesis. An observer-based fault diagnosis method is formulated to detect and isolate the faults under consideration. For each fault, a fault detection residual is used to indicate the occurrence of the fault, and a set of fault isolation residuals are developed to determine the type and the location of the fault. Each isolation residual is designed based on a particular fault scenario under consideration. The FDI design is achieved by utilizing certain particular structures of the benchmarked model without estimation the wind speed or the aerodynamics of the turbine. Simulation results are presented to illustrate the effectiveness of the developed method.
Department or Program
Department of Electrical Engineering
Year Degree Awarded
Copyright 2011, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.