A Gradient Correction-based Decentralized Optimal Var/Volt Adaptive Fault-Tolerant Control Method for Wind Farms

This paper proposes a gradient correction-based decentralized optimal Var/Volt adaptive fault-tolerant control (GC-AFTC) method for wind farms (WFs). The aim is to ensure the decentralized optimal Var/Volt operation of wind turbines (WTs) inside a radial topology WF even under the fault condition due to several WTs being tripped. When some faults occur and several WTs are tripped, the decentralized optimal Var/Volt operation of WF still can be maintained without the reconstruction of the incidence matrix and any quantitative fault information of the fault WTs. To achieve the fault-tolerant optimal Var/Volt control, the impact of random off-grid events (ROEs) on the WF collection system is analyzed at the farm level. Subsequently, a unified GC method is derived based on the gradient projection (GP) method. The transient Var related to the ROE is incorporated into the GC method to enhance the capability of decentralized online applications of WTs. Additionally, to improve the dynamic performance of WTs, a personalized step coefficient is designed within a stable range, which is utilized in each local WT controller during the GC-AFTC period. Case studies show that the proposed method can drive the WF system towards a better operating point only with local measurements in normal and fault conditions.