Evaluation and Adjustment Based on Broad Learning System for Small Signal Stability Analysis of Power System Integrated with Wind Farms

This paper primarily focuses on evaluation and adjustment methods for small signal analysis with wind farm integration. The connection of wind farms may affect the original oscillation modes of the system, and may also introduce new oscillation modes. To ensure the security and stability of the system, it is necessary to assess the running state of the system and evaluate the development direction of the small signal stability on the premise of taking volatility into account. It is also important to fast adjust the system when an unstable state happened. The method in this paper provides a new way to accurately evaluate and adjust the stability of the system integrated with wind farms. Broad learning system (BLS) is a non-fixed model and has the potential of incremental learning. A BLS model is trained to evaluate the damping ratio sensitivity in different operating states. The adjustment part will use the optimal adjustment model and damping ratio sensitivity to adjust unstable systems. The superior performance of the model proposed in this paper is tested by the case study based on the IEEE 3-machine, 9-bus, and the New-England 10-machine, 39-bus. The results demonstrate the feasibility, validity, and effectiveness of the method proposed in this article.