The　production　process　of　ground　granulated　blast　furnace　slag　(GGBS)　aims　to　produce　products　of　the　best　grade　and　the　highest　yields.　However,　grade　and　yields　are　two　competing　objectives　which　can　not　be　optimized　at　the　same　time　by　one　single　solution.　Meanwhile,　the　production　process　is　a　multivariable　strong　coupling　complicated　nonlinear　system.　It　is　hard　to　establish　the　accurate　mechanism　model　of　this　system.　Considering　above　problems,　we　formulate　the　GGBS　production　process　as　an　multiobjective　optimization　problem,　introduce　a　least　square　support　vector　machine　method　based　on　particle　swarm　optimization　to　build　the　data-based　system　model　and　solve　the　corresponding　multiobjective　optimization　problem　by　several　multiobjective　optimization　evolutionary　algorithms.　Simulation　example　is　presented　to　illustrate　the　performance　of　the　presented　multiobjective　optimization　scheme　in　GGBS　production　process.