Advanced control for large scale non-conventional power systems

• Main Author: Wu, Si
• Other Authors: Wang Youyi
• Format: Theses and Dissertations
• Language: English
• Published: 2012
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electric power
• Online Access: https://hdl.handle.net/10356/50762
• Institution: Nanyang Technological University

This thesis concerns the applications of advanced controls for large scale nonconventional power systems. The so-called non-conventional power system is defined as the power system with grid-connected renewable power generation sources in this thesis. Among all kinds of renewable power generation sources, the installed capacity of wind power is the largest with the growing over 25% per year. Thus, it is reasonable and significant to set wind power as a representative of the renewable power generation sources in the non-conventional power system. In this work, researches are concentrated on the self-control of wind turbine power generation system in non-conventional power systems and the global control of the entire non-conventional power system. From a Wind Energy Conversion System’s (WECS) point of view, it always desirable to capture as much power as possible. This depends on both that the WECS is able to obtain accurate and real-time wind information and efficient and accurate controls are employed in the WECS control systems. For example, for distributed WECS or grid-connected WECS in the case of normal operation of the power system, WECSs main concern is its own control, including wind turbine Maximum Power Point Tracking (MPPT), real-time wind speed estimation, turbine pitch system control and sensorless control of WECS, etc. In the controls of variable-speed variable-pitch Wind Turbine Power Generation System (WTPGS), the turbine shaft rotational speed should be controlled optimally with respect to the variable wind speed in order to achieve MPPT. A precise real-time wind speed estimation method and sensorless control for variable-speed variable-pitch WTPGS are proposed in this thesis. The wind speed estimation is realized by an Extreme Learning Machine (ELM) based nonlinear input-output mapping neural network. A specific design characteristic of the wind turbine is used for improving the mapping accuracy while considering the control of turbine pitch angle system. The proposed wind speed estimation is established by using the information of the wind mechanical torque and the turbine rotational speed which are estimated by a two-loop linear observer. The estimated wind speed is then used to determine the optimal rotational speed and pitch angle commands.