Social　emotional　optimization　algorithm　(SEOA)　is　a　new　algorithm　inspired　by　the　human　social　behavior.　In　this　paper,　a　new　variant,　named　discrete　social　emotional　optimization　algorithm　with　Lattice,　is　designed　to　solve　the　structural　optimization　of　Lennard-Jones　(LJ)　clusters.　Lattice　technique　is　a　strategy　to　optimize　the　atoms　on　its　possible　positions,　with　this　manner,　the　total　search　space　of　Lennard-Jones　clusters　problem　turns　to　several　discrete　points,　thus,　the　discrete　SEOA　is　employed　to　solve　it.　To　improve　the　local　search　capability,　L-BFGS　is　also　employed　to　enhance　the　exploitation　capability.　Simulation　results　show　the　validity.