Multi-objective　optimization　is　an　important　and　challenging　topic　in　the　field　of　industrial　design　and　scientific　research.　Evolutionary　algorithm　is　a　population-based　meta-heuristic　technique　to　effectively　solve　Multi-objective　Optimization　Problem　(MOP).　In　this　paper,　a　novel　EA　is　proposed,　which　applied　the　construction　strategy　of　the　elitist　population　based　on　spacial　grid.　In　this　strategy,　firstly,　a　fast　obtaining　Pareto　set　approach　with　less　computation　cost　is　employed;　then　we　filter　Pareto　set　with　the　grid　with　the　fixed　side　length　to　keep　the　diversity　of　solutions.　Experimental　results　on　test　problems　show　that　the　GSEA　we　proposed　improves　time　performance　significantly,　and　is　able　to　find　solutions　with　good　diversity　and　being　nearer　the　true　Pareto-optimal　front　compared　to　the　known　NSGA-II,　SPEA2　and　Ε-MOEA.　©　2011　IEEE.