Gradient search optimized interval type-2 fuzzy TID approach for stability enhancement of an islanded AC microgrid

• Designed a hybrid microgrid in MATLAB Simulink with wind, solar, fuel cells, and microturbines for renewable integration. • Developed an IT2-FTID controller to regulate frequency deviations under step loads, wind uncertainties, and solar variations. • Compared IT2-FTID with type-1 fuzzy PID and conventional PID to evaluate control effectiveness in microgrid stability. • Proposed a gradient-based algorithm (GBA) to optimize IT2-FTID gains, outperforming PSO and GA in controller tuning. • Validated IT2-FTID robustness through extensive sensitivity analysis, ensuring reliable performance under system variations. The islanded AC microgrid incorporates with low inertia and huge uncertainties, for which frequency instability issues are occurred under load dynamics and solar/wind uncertainties. This research paper proposed a resilient interval type-2 fuzzy tilt integral derivative (IT2-FTID) control strategy for obtaining frequency stability in the microgrid under various uncertainties. The model of an AC microgrid is developed with integrating multiple power-generating units based on renewable energy sources like solar and wind. These units, having capacities in the fractional megawatt range, form what is termed a microgrid. The microgrid's performance is influenced by significant load dynamics and uncertainties in wind speed and solar intensity, particularly affecting system frequency. To mitigate unwanted frequency fluctuations and maintain nominal frequency during various electrical disturbances, the resilient IT2-FTID controller operates as a secondary frequency control loop in the microgrid. Additionally, a novel Gradient-based Algorithm (GBA) with an effective objective function is applied to optimize the performance of the IT2-FTID controller. The effectiveness of the IT2-FTID approach is compared against basic type-1 fuzzy and conventional PID controllers, showing a frequency stability improvement of 111.76% and 205.88%, respectively. Finally, the convergence and optimization capabilities of the GBA are compared to Genetic Algorithm (GA) and standard Particle Swarm Optimization (PSO) for demonstrating superior performance.