Please use this identifier to cite or link to this item:
|Title:||Advanced control of wind power systems||Authors:||Dang, Dinh Quy||Keywords:||DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Control engineering
DRNTU::Engineering::Electrical and electronic engineering::Electric power::Auxiliaries, applications and electric industries
|Issue Date:||2012||Source:||Dang, D. Q. (2012). Advanced control of wind power systems. Doctoral thesis, Nanyang Technological University, Singapore.||Abstract:||Wind power system control especially for variable speed operation of wind turbine has been carried out in several decades; and the research activities are still on-going with more concern recently. The reason is that a variable speed wind power system produces higher electrical energy and power efficiency than a conventional wind power plant. However, the variance of rotor speed leads to the changes of working conditions in accordance with stochastic wind profiles. The different control objectives for each working condition are also addressed as a problem of controller design. Beyond the multiple operating regions, wind power system control has to deal with the nonlinearity of wind torque and the stochastic behavior of wind. Hence, the control system is further complicated. The wind power system configuration in this study includes a variable speed wind turbine driving a permanent magnet synchronous generator. The generator is connected to a full-scale back-to-back IGBT converter which is connected to a local grid. This wind turbine configuration is regulated by four individual controllers, which are speed controller of turbine drive-train; generator d-q axis current controllers; dc-bus voltage control of dc-link; and grid side d-q axis currents (also known as grid active and reactive power) control. Among these controllers, the classical PI (proportional integrator) controller has been well applied to d-q axis currents of both generator and grid side from the literature so that it is not a main study in this thesis. The control algorithm for aerodynamics drive-train and dc-link voltage is therefore of the research interest. Through the simulation and experimental results, it is concluded that wind power system control problems are well resolved by proposed algorithms. There are several methods which are applied for controlling the aerodynamic drive train loop based on model-free, linear to nonlinear models. Model-free fuzzy logic controller (FLC) acts as a human behavior to find appropriate control input based on the knowledge of measurement signals and rules. FLC provides a simple and good optimal power tracking at partial-load region; while the nonlinear PID controller with variable gains gives a simple, stable control scheme for different operating points. The simplicity of these two methods makes it popular and suitable for small and medium rating wind power systems. The wide load conditions of wind power system control can be obtained by linear Model Predictive Control (MPC). At partial load working region, a reference tracking quadratic cost function is formulated. Cut-in and full load conditions are considered as lower and upper regions which are integrated as constraints to the partial load condition. The cost function with constraints is then minimized in every sampling time to find optimal control input updates to the plant. Moreover, linearization presentation of nonlinear system always creates mismatch between model and plant hence produces a reference tracking offset. The robust Disturbance Models of Predictive Control (DMPC – modification method from linear MPC) is based on uncertainty linear model which adds a disturbance to the output and an appropriate observer designed for estimation of states to cancel the steady state errors. Last control algorithm is an extended method of linear MPC, the Nonlinear MPC (NMPC) designs directly from wind power nonlinear model. All the properties of linear MPC are inherited in the nonlinear version with better quality. A real-time process which includes multiple shooting method, partially-reduced sequential quadratic programming and real-time iteration scheme is applied to realize NMPC to wind power application. The final development in wind power control is equipped with robust DMPC method for a cluster of few wind turbines in connection to the grid. The robust MPC controller for dc-link voltage in cascaded with PI controller of the generator side d-q axis currents is proposed. The result of wind turbine cluster can be extended to a large scale wind farm. Fault-ride through capability of wind farm control is also analyzed.||URI:||http://hdl.handle.net/10356/50720||Fulltext Permission:||open||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Theses|
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.