Yuhao Zhou

ORCID Identifier(s)


Graduation Semester and Year




Document Type


Degree Name

Doctor of Philosophy in Electrical Engineering


Electrical Engineering

First Advisor

Wei-Jen Lee


With high penetration level of the wind power energy in the power grid, it is necessary to develop robust dynamic wind farm models for stability analysis under different scenarios. However, one single wind generator (WG) contains many parameter and control loops, it’s impractical to apply such detailed high-order model of individual wind turbine for dynamic stability analysis due to its computational burden and availability of the data. Therefore, there’s a need for deriving the dynamic equivalent model (DEM) for the wind farm. In the real world, a wind farm may contain different types of WGs coming from different manufactures/vendors with crowbar protection schemes, this makes dynamic equivalent modeling process very challenging and difficult. Western Electricity Coordinating Council (WECC) proposed 4 types of WG’s generic models. In this article-based dissertation, the WECC’s generic models are applied, and the proposed detailed procedures of deriving the DEM for power grids’ planning, operation and post-fault analysis are presented in 4 published research papers. In the first paper, the single-machine DEM for a wind farm containing type-3 doubly fed induction generators (DFIGs) coming from the same manufacture/vendor was conducted. The parameter identification for the derived DEM was performed by adopting the novel heuristic optimization algorithm, which aims at minimizing the differences of the dynamic responses measured at the point of interconnection (POI) between the detailed wind farm model and the derived DEM. Moreover, to make it more practical and robust considering the realistic scenario that only portion of the WGs are online, the proportion estimation analysis was adopted, which aims at adaptively modifying the ratings of the DEM to improve its robustness. To further improve the feasibility of the proposed method, a multistage dynamic equivalent modeling method for a wind farm consists of different types of WGs coming from different manufactures/vendors was presented in the second paper. Sensitivity analysis was applied for choosing the key parameters to improve the speed of the model validation process. Besides, the wake effect model was considered regarding the robustness of the derived DEM. Since type-3 WGs (DFIG) usually adopt the crowbar protection scheme regarding the low voltage ride through (LVRT) requirement and the overcurrent protection of the rotor in the generator, it will affect the dynamic response of the DFIG. However, there’s no crowbar circuit component in WECC’s generic models. To enhance the applicability of the model, the third and fourth papers in this dissertation add the crowbar circuit component into the WECC’s type-3 generic model to derive the DEM of the DFIG wind farms considering the crowbar protection. In the third paper, the wind speed distribution inside the wind farm is assumed to be known, and thus an improved multi-machine DEM was developed to improve the accuracy and robustness. In the fourth paper, the DEM was developed for the post-fault analysis of the wind farm based on the crowbar circuit’s triggering waveforms if the previous assumption about the wind speed distribution inside the wind farm was unknown.


Wind generation, Wind farm modeling, Dynamic equivalent model


Electrical and Computer Engineering | Engineering


Degree granted by The University of Texas at Arlington