Supervisory voltage control scheme for grid-connected wind farms.

摘要

Modern variable speed wind turbines utilize power electronic converters for the grid connection requirement and to improve performance. Most commonly used converters enable the wind turbines to maintain the required power factor (power factor control) or voltage (local voltage control) at the terminals. However, in many wind farm applications there is a need to control voltage at a specified remote location, which may require the installation of additional compensating devices (transformer tap changers, switched-capacitors, SVCs, etc.) to meet local power quality conditions. This thesis proposes a supervisory control scheme that uses the individual wind turbines to regulate voltage at the required location, i.e., point of common coupling. The proposed approach considers that each turbine may have somewhat different instantaneous wind speeds and real power outputs, and therefore different amounts of reactive power available for achieving the main control objectives. The operating limits of each turbine are also taken into account to ensure that all power electronic converters operate in the allowable region. Since the proposed supervisory scheme is general and can work with different controllers, we investigate several controllers in this thesis. The problem of control design is formulated as a linear matrix inequality. An innovative cost-guaranteed linear-quadratic-regulator-based controller with an observer is proposed and tuned for a range of operating conditions.; In this thesis, we apply the proposed supervisory voltage control methodology to a candidate wind farm site on Vancouver Island, BC, Canada, made available through collaboration with Powertech Labs Inc. We have developed a detailed model of the system, using three 3.6 MW wind turbines, to carry out the simulation studies. The proposed control solution is compared with traditional approaches and shown to be very effective during load disturbances and faults. The proposed methodology is also flexible and readily applicable to larger wind farms of different configurations.