To　avoid　the　increasing　of　the　power　loss　and　the　declining　quality　caused　by　the　imbalance　of　reactive　power　in　power　system,　a　method　based　on　the　co-evolution　of　Simulated　Annealing　algorithm　(SA)　and　Particle　Swarm　Optimization　(PSO)　for　reactive　power　optimization　is　established.　PSO　is　a　evolutionary　computational　method　that　optimizes　a　problem　fast　and　effectively　and　it　is　suitable　for　handling　real-value　problem.　But　PSO　algorithm　also　has　some　limitations　such　as　premature　convergence,　which　causes　the　bad　accuracy　of　convergence.　Some　expansions　and　corrections　including　inertia　weight　and　Simulated　Annealing　algorithm　(SA)　are　introduced　in　this　paper　to　improve　the　basic　particle　swarm　optimization　algorithm.　Simulated　annealing　algorithm　with　the　weight　coefficient　combined　with　particle　swarm　optimization　(SA-WPSO)　is　proposed　to　solve　reactive　power　optimization　problem.　Compared　with　the　algorithms　such　as　PSO　and　SA-PSO,　SA-WPSO　is　better　for　global　convergence　and　higher　accuracy　of　reactive　power　optimization　by　using　IEEE-10　bus　system　as　a　model　for　the　simulation.　©　2012　Chinese　Assoc　of　Automati.