Hybridizing gaining–sharing knowledge and differential evolution for large-scale power system economic dispatch problems

Abstract Economic dispatch (ED) of thermal power units is significant for optimal generation operation efficiency systems. It a typical nonconvex and nonlinear optimization problem with many local extrema when considering the valve-point effects, especially large-scale Considering that differential evolution (DE) efficient in locating global region, while gain-sharing knowledge-based algorithm (GSK) effective refining solutions, this study presents new hybrid method, namely GSK-DE, to integrate advantages both algorithms solving ED problems. We design dual-population framework which population randomly divided into two equal subpopulations each iteration. One subpopulation performs GSK, other executes DE. Then, updated individuals these are combined generate population. In such manner, exploration exploitation harmonized well improve searching efficiency. The proposed GSK-DE applied six cases, including 15, 38, 40, 110, 120, 330 units. Simulation results demonstrate gives full play superiorities GSK DE effectively. possesses quicker convergence rate obtain higher quality schemes greater robustness. Moreover, effect size also examined.