AI Summary & Key Takeaways
AI InsightSimulation results show that iDNA-MOS converges faster than its competitors in tests such as Sphere and Ackley.
📌 3 Key Takeaways
Abstract
In the management of modern smart grids, the detection of non-technical losses (loads caused by unauthorized use or cyber-attacks) and the simultaneous minimization of production costs constitute a complex optimization problem. Classical meta-heuristic algorithms typically model the system under ideal conditions and disregard physical anomalies on the grid. This study proposes a new iDNA-Based Mosquito Optimization Strategy (iDNA-MOS), inspired by the ability of mosquitoes in biological research to analyze species based on blood samples taken from hosts (iDNA - invertebrate DNA). The proposed method combines standard mosquito movement mechanisms (Lévy flight) with a unique bio-diagnostic operator that diagnoses network imbalances. The performance of iDNA-MOS was evaluated on the IEEE 30-Bus test system, which includes five different mathematical benchmark functions and 25 MW of hidden theft load. Simulation results show that iDNA-MOS converges faster than its competitors in tests such as Sphere and Ackley. On the IEEE 30-Bus system, iDNA-MOS demonstrated statistically significant (p < 0.05, Wilcoxon Rank-Sum) superior performance compared to the standard Mosquito Algorithm in the literature (Lit-MFO: 942.24 $/h), with an hourly cost of 901.87 $/h. Although it produced results similar to the industry standard PSO in terms of cost (p ≈ 0.95), iDNA-MOS was the algorithm that most accurately compensated for the hidden load by reaching a total generation capacity of 312.13 MW, which is critical for system reliability.