Studies On Risk Warning For Trip-off Of Wind Turbine And Anti-off-grid Strategies Under Multiple Scenarios For Clustering Wind Farms

As the most mature and commercialized power generation technology based on renewable energies, wind power generation has become an important solution for China to tackle energy problems and climate change. However, under the development mode of the highly-concentrated wind power base, wind farms within a certain geographical range are highly susceptible to the same electrical disturbance or weather event. The frequent occurrences of wind power off-grid accidents in recent years have revealed many deficiencies on operation stability and fault ride through ability of wind farm. If early warning information and effective anti-off-grid strategies can be provided before accident occurrence or deterioration, risk and/or loss due to trip-off of wind turbines can be greatly reduced. Therefore, it is an urgent task to establish systematic framework and theories on wind turbine trip-off accident early warning and anti-off-grid solutions. Therefore, scholars and specialists in the field of academy and industry have put forward many elaborate ideas and solutions to enhance the perceptual ability of the wind farm transient survivability and wind turbine fault ride through ability. However, it is still worth studying on off-grid risk warning concerning the discrete and difference between wind turbines, impact of severe wind conditions on wind farm, performance improvement of risk warning using data mining techniques, and coordination between operation margin of protection and low voltage ride-through capability, etc. Therefore, this thesis focuses on the operation and control issues of large-scale concentrated wind farms, and its emphasis is placed on risk warning methods for wind turbine trip-off accidents and anti-off-grid strategies under multiple scenarios.In the aspect of off-grid risk warning method and index techniques, the following work have been fulfilled in this thesis:In order to achieve the quantitative evaluation on voltage sag risk level and consequences, together with location of disturbance sources, a probabilistic theory based wind farm voltage-dip risk assessment methodology and a series of new indexes capable of indicating voltage-dip risk level are proposed, and a wind farm voltage-dip evaluation program is implemented based on Digsilent. By means of this work, a novel method can be provided for optimal siting of grid-connection point of wind farm and improving the ability of wind farm resistive to voltage fluctuation. In order to reveal the impact of various wind conditions on wind farm operation and wind turbine load-bearing, it is necessary to quantitatively simulate the standard wind conditions in view of wake effect, average wind speed and turbulence, etc. The synthetic simulation technique of wind operation taking the impact of wind turbine bearing-load and dynamic wind condition into account, the constructing method of interactive simulation platform are studied. Furthermore, corresponding synthetic simulation platform is established. By means of this work, a novel countermeasure can be provided for analyzing valuable scenarios such as wind turbine crash due to severe wind condition, sudden power decrease, and their impact on the power system connected by this wind farm. In order to satisfy requirements of rapidity and accuracy for real-time risk warning of wind turbine off-grid, based on classification tree (CA) algorithm in decision tree theory (DT), by means of utilizing the excellent characteristics of DT-type algorithm including high transparency and good output compatibility, the key impact factors of wind turbine off-grid can be revealed. Furthermore, wind turbine off-grid risk measurement under various contingencies can be judged in advanced and corresponding disturbance measure indexes can be output. The achievement of this work can provide intuitive decision reference for dispatcher of power systems.In the aspect of anti-off-grid techniques, the following work have been fulfilled in this thesis:In order to improve power generation reliability of wind farm without off-grid under severe wind conditions, based on impact analysis of wind conditions on turbine load, wind turbine safety operation constraints are constructed, and a novel cohesive control strategy of wind turbine based on damage equivalent load tracing is proposed. By means of this work, a novel method can be provided for more effectively utilizing wind turbine load bearing capacity and enhancing anti-oflf-grid ability of wind turbine. For the purpose of overcoming the shortage of MCR-type SVC control on regulating speed in response to voltage very-fast recovery of wind farm and therefore effectively controlling the transient over-voltage level of wind farm due to fault removal, a new MCR-type SVC overvoltage emergency control strategy based on regression tree (RT) in DT is proposed. By means of this work, a novel countermeasure can be provided for lowering the off-grid risk of wind farm due to voltage overshoot. In order to deeply explore the low voltage ride-through ability of wind farm, based on the analysis of wind turbine having low-voltage ride-through ability still failing to ride through under some voltage disturbance scenarios, a new wind turbine active control strategy and its intelligent upgrade solution are put forward. By means of this work, a novel thought can be provided for further improving low-voltage ride-through ability of wind farm.