摘要：Computer technology provides new possibilitiesfor handling the many-objective optimal power flow (MOOPF)problems with high-dimension and non-differentiability. As oneof typical intelligent algorithms, the novel multi-objectiveartificial fish swarm algorithm (NMAFSA) is proposed to solvethe MOOPF problems and realize the economical operation ofpower systems. The NMAFSA algorithm, which combines withoptimal solution guidance (OSG) principle and non-inferiorretention (NIR) mechanism, is effective to reduce the fuel cost,emission and power loss. Compared with the representativemany-objective particle swarm optimization (MPSO) andnon-dominated sorting genetic algorithm-Ⅱ (NSGA-Ⅱ), thesuperiority and adaptability of presented NMAFSA algorithmare validated. Six simulation trials are carried out on MATLABsoftware, including the dual-objective and triple-objectiveoptimizations on three different scale power systems. Detailedresults demonstrate that the suggested NMAFSA algorithm withstable-operation and fast-convergence has great potential to dealwith the MOOPF problems more efficiently. Furthermore, thegeneration distance (GD) index also quantitatively proves thatthe NMAFSA algorithm can obtain the well-distributed Paretofront (PF).
关键词：Artificial fish swarm algorithm; Optimal power flow; Computer technology; Generation distance
