Intelligent Particle Filtering State Observer for Stability Assessment in Solar-Wind Penetrated Microgrids

This paper presents an advanced Intelligent Particle Filtering State Observer (PFSO) for real-time voltage and frequency stability assessment in microgrids integrated with high-penetration solar and wind energy sources. The proposed method leverages the robustness of PFSO to address the nonlinear, stochastic, and dynamic behaviors inherent in renewable-based distributed generation systems. A comprehensive state-space model of the microgrid is developed, and the PFSO is employed to estimate unmeasurable or noisy system states in the presence of process and measurement uncertainties. The proposed method was validated in MATLAB/Simulink across three scenarios: normal operation, sudden power mismatch, and periodic load disturbance. Quantitative results demonstrate that the PFSO maintains high estimation accuracy, with Root Mean Square Error (RMSE) values consistently below 0.0095 per unit (p.u.) and Mean Absolute Error (MAE) under 0.0073 p.u. for both voltage and frequency states. The maximum estimation error remained below 0.020 p.u., confirming strong robustness under transient conditions. Furthermore, a binary classification analysis of system stability, using a 0.95 p.u. threshold achieved 97.4% accuracy, 95.9% precision, and an F1-score of 96.5% across all cases. The findings validate the effectiveness of the proposed PFSO as a reliable tool for dynamic state estimation and early instability detection in smart microgrid environments.