Abstract

The susceptibility of modern interconnected microgrid (MG) systems to undesirable frequency oscillations represents a critical issue resulting from the intermittency properties of renewable energy generation systems (RESs). Thence, extensive efforts are needed for developing more robust frequency regulators. Maintaining required frequency regulation using load frequency controllers (LFCs) can play a vital role in modern interconnected multi-MG systems with high levels of RESs penetration. Moreover, wide concerns exist about taking advantage of installed electric vehicle (EV) battery energy storage devices in regulating the frequency of MG systems. Therefore, this paper proposes an improved optimal fractional order (FO) LFC technique for developing robust multi-MG systems. The proposed technique uses two cascaded control loops, wherein the one plus proportional double derivative with filter (1+PDDF) in the outer loop and FO proportional integral tilt (FOPIT) control technique. To avoid complex control design procedures and parameter tuning, the recent powerful artificial hummingbird optimizer algorithm (AHA) is proposed to simultaneously optimize proposed control parameters in various interconnected multi-MG systems. The AHA optimizer outputs optimal parameters of the proposed control to achieve the best system response and stability. The proposed 1+PDDF/FOPIT control scheme yields a high rejection rate of existing disturbances resulting from uncertainties in RESs and/or loads, which enhances frequency stability and robustly mitigates frequency oscillations. Several comparisons and evaluations are provided for the proposed 1+PDDF/FOPIT control scheme and AHA optimizer with featured existing schemes in literature to prove several superiority indices of system response. For instance, the integral squared error (ISE) comparison at step load scenario for the proposed 1+PDDF/FOPIT controller was 0.31%, 0.29%, 1.25%, and 4.12% of the obtained ISE values under the studied PIT, FOPIT, 1+PDDF, and 1+PIDF/FOPID LFC methods from literature, respectively.