Particle　Swarm　Optimization　(PSO)　has　proved　to　be　an　effective　global　optimization　in　recent　years.　However,　PSO　still　suffers　from　the　prematurity　to　local　optima.　In　order　to　solve　this　disadvantage,　researches　have　carried　out　by　combination　with　other　optimizers.　In　recent　years,　a　local　optimization　called　Extremal　Optimization　(EO)　has　been　introduced　into　PSO　and　gain　improvements.　Although,　the　combination　of　PSO　with　EO　would　bring　severe　computation　overhead　result　in　the　simple　way　they　combined.　In　this　paper,　an　adaptive　hybrid　combined　PSO　(AHPSO-EO)　is　proposed.　It　can　improve　the　computation　effectively　by　an　adaptive　way.　The　experimental　results　on　benchmark　functions　reveal　that　the　adaptive　PSO　combined　with　EO　accelerates　the　convergence　and　improves　the　performance　of　proposed　algorithm.　©　2012　IEEE.