To mitigate the frequently over‐threshold subsynchronous oscillation (FOSSO) phenomenon which occurred in Hulunbuir power plant in the northeast of China, a novel equipment based on voltage source converter (VSC), named subsynchronous oscillation dynamic suppressor (SSODS) is developed and installed. In this study, the root cause of FOSSO is revealed by analysing the response characteristics of the system. Then the control strategy of the developed SSODS is presented. Damping characteristics of SSODS are analysed by the complex torque coefficient method. To verify the effectiveness of the control strategy and the physical controller of VSC, as well as provide initially optimised controller parameters for the field tests, real‐time closed‐loop tests are carried out. A novel method of torsional modal frequency measure for onsite commissioning test is proposed. The operation and field verification test of SSODS is introduced, and test results demonstrate that SSODS can mitigate the new SSO phenomenon dramatically, proving the validity of SSODS.
This study offers a modular isolated grid-connected DC/DC medium-voltage DC aggregation converter to support offshore full DC wind farms’ need for lightweight and highly efficient power aggregation and transmission. The converter can simultaneously have a smaller transformer size and lower switching frequency during operation through the dual-voltage stabilization three-loop control strategy and phase-shift modulation strategy, which greatly reduces the space occupied by the converter and lowers the switching loss, Additionally, the use of a two-level structure at a lower switching frequency has lower loss, which effectively reduces the cost of the power device compared with the commonly used three-level converter. The input series output series connection between the converter sub-modules effectively lowers the voltage stress on each power switching device and facilitates expansion into a multi-module structure, expanding its application in high-voltage and large-capacity environments. This study analyzes the two working modes of the DC/DC converter and its control approach, in addition to providing a detailed introduction to the application scenarios of this converter. Ultimately, the efficacy and practicability of the suggested topology and control scheme are confirmed by simulations and experiments.
Large-scale wind-farm clusters are characterized by intense power fluctuations, various turbine types and time-varying reactive power margins, along with conflicting goals of voltage security and network loss reduction. To address these issues, this paper proposes a hierarchical coordinated reactive power optimization strategy based on an Environment-Aware Particle Swarm Optimization (EAPSO) algorithm. In this study, “environment-aware” specifically refers to operating-condition awareness, including low-load degree, spatial power dispersion, and output imbalance degree of wind turbines. A three-layer framework consisting of cluster level, wind-farm level and unit level is established to coordinate wind-farm reactive power commands, SVG outputs, transformer tap positions, and turbine-level reactive power allocation. A comprehensive disturbance index is constructed from the low-load degree, spatial power dispersion, and output imbalance degree and is then mapped to the initialization strategy, learning factors, and mutation operation of EAPSO. A 24 h offline simulation is conducted using historical operating profiles of a practical large-scale wind-farm cluster in South China. The results show that the proposed strategy reduces the average absolute PCC voltage deviation from 0.013331 p.u. before optimization to 0.004120 p.u. it achieves an average network-loss reduction rate of 3.55%. Compared with Method 1, the transformer tap operation times decrease from 139 to 92, and the cumulative SVG reactive output decreases from 1077.32 Mvar to 208.39 Mvar.
Grid-following converters are prone to transient synchronization instability during grid faults, which may lead to the disconnection of renewable energy devices and threaten the security and stability of power systems. To address this issue, this paper first establishes a second-order nonlinear reduced-order model centered on phase-locked loop dynamics and constructs an energy function to analyze the instability mechanism, revealing the decisive role of equivalent damping in the energy dissipation process. On this basis, an adaptive damping stabilization control strategy based on q-axis voltage feedback is proposed. By introducing the amplitude of the q-axis voltage into the feedback loop to reshape the damping path, the strategy can adaptively adjust the damping according to the fault severity. This control strategy significantly enhances the synchronization stability of grid-following converters: it suppresses oscillations and achieves fault ride-through during transient instability, while limiting the divergence of the power angle during static instability to buy time for fault clearing. Finally, hardware-in-the-loop experiments validate the effectiveness of the proposed strategy.
Enhancing the accuracy of short-term wind power forecasting helps mitigate the adverse impacts of prediction errors on grid dispatch. Wind power exhibits a significantly nonlinear dependence on multiple influencing factors. However, existing methods struggle to effectively resolve multi-dimensional feature redundancy and multi-scale non-stationary evolutionary characteristics inherent in far-offshore wind power forecasting tasks. This leads to bottlenecks such as insufficient feature discriminability and temporal dependency focus shift under complex marine environments, ultimately limiting further improvements in prediction accuracy. To address these challenges, this paper proposes a federated learning-based adaptive multi-head attention model for wind power forecasting (Fed-AMHA). The proposed framework operates as follows: First, each wind farm client utilizes a Bidirectional Long Short-Term Memory (BiLSTM) network to model input sequences bidirectionally, capturing long-term temporal dependencies. Subsequently, linear projection and parallel one-dimensional convolution operations are introduced to mine multi-scale local temporal features from each time step and its neighborhood. Building upon this, channel attention and multi-head temporal feature attention mechanisms are stacked. The model adaptively adjusts the weights of different time slices and feature channels by learning the importance of each channel to the forecasting task. The central server then aggregates the model parameters uploaded by the clients via averaging, enabling cross-site collaborative training without directly sharing raw data. Simulation results based on public datasets and actual wind farm data under various short-term forecasting scenarios demonstrate that the proposed model consistently achieves lower prediction errors and superior stability compared to existing forecasting models under identical settings.