Disturbed security-constrained and time-variant optimal power flow for dynamic power system based on chaotic-genetic-centroid puffin optimization

Although the Security-Constrained Optimal Power Flow (SCOPF) model provides an effective model for simulating fault scenarios in power systems, it often overlooks bus-level fluctuations. Therefore, this paper introduces the Disturbed Security-Constrained Optimal Power Flow (D-SCOPF) to model the instability of power systems, leading to the development of a disturbed topology varying (DTV) test system. Additionally, to simulate the dynamic and time-varying properties of power systems, the fluctuations in electric vehicle load are incorporated into the system, resulting in the construction of a Time-Variant (TV) test system. These enhancements improve model realism but also significantly increase computational complexity. As a result, finding feasible solutions in such a complex model becomes a substantial challenge. Considering the high-dimensional and multi-constraint properties of the OPF problem, this paper proposes a novel Chaotic-Genetic-Centroid Puffin Optimization (CGC-PO) algorithm based on Arctic Puffin Optimization (APO). CGC-PO aims to specifically improve APO’s exploration, exploitation, and population merging processes through centroid opposition-based learning, normalized chaotic local search, and genetic hybrid incorporation. To validate the feasibility of the proposed algorithm, a series of tests were first conducted on benchmark functions, followed by quantitative, convergence, and statistical analyses. Besides, different modified test systems were constructed based on the IEEE 30-bus, IEEE 57-bus, IEEE 118-bus and Illinois 200-bus systems. Both single-objective and multi-objective optimization functions were evaluated on these test systems. By comparing the results with those obtained from several well-known optimization algorithms, including TLBO, PSO, PLO, SMA, HGS, MGO, ALA, MSO and APO, the effectiveness, superiority, and robustness of the proposed CGC-PO algorithm are demonstrated.