Outlier-Resistant Recursive State Estimation for Renewable-Electricity-Generation-Based Microgrids

Chapter 2 : This chapter addresses the state estimation (SE) problems for renewable-electricity-generation (REG)-based microgrids, where the measurement outliers are also considered. Different from the traditional weighted least square or Kalman filter (KF)-based methods, which lack robustness to outliers and fail to adapt to microgrid dynamics, the chapter proposes a novel outlier-resistant recursive SE algorithm. Key innovations include embedding a saturation function to constrain the innovation term, mitigating adverse impacts from outlier-contaminated data without requiring prior knowledge of outlier characteristics (e.g. type, frequency, or probability). The algorithm also guarantees an upper bound on the estimation error covariance, which is minimized by optimally designing the estimator gain. Based on three simulation scenarios (gross errors, large measurement noises, and random outliers) on an islanded microgrid with two REG units, it can be found that the proposed approach outperforms conventional state estimation methods in estimation accuracy and robustness. In summary, this work provides a practical solution for real-time monitoring, security assessment, and situational awareness of REG-based microgrids, which is useful in enhancing the reliable operation in both grid-connected and islanded modes.