Genetic　algorithm　(GA)　is　a　predominant　optimization　algorithm　with　the　advantages　of　a　wide　application　range,　high　success　rate　and　strong　adaptability.　However,　GAs　often　suffer　from　the　issue　of　falling　into　local　optima　because　the　excessive　population　fitness　value　generated　during　the　global　search　process　reduces　the　population　diversity　of　GA.　To　overcome　this　issue,　this　paper　proposes　an　adaptive　genetic　algorithm　based　on　simulated　annealing　strategy　(SA-AGA).　Primarily,　a　simulated　annealing　strategy　is　designed　to　achieve　the　local　search　in　SA-AGA.　The　simulated　annealing　strategy　utilizes　the　fitness　of　the　population　to　evaluate　the　new　solution　and　randomly　searches　for　the　preferable　solution　around　the　current　solution　during　a　stated　cooling　process.　It　enables　SA-AGA　to　effectively　jump　out　of　the　local　optimum.　Then,　two　adaptive　genetic　operators　are　developed　in　SA-AGA　to　achieve　the　crossover　operator　and　the　mutation　operator.　These　adaptive　genetic　operators　enhance　the　genetic　probability　adaptively　when　encountering　excessive　fitness,　which　further　improves　the　population　diversity　of　SA-AGA　and　ensures　stable　and　rapid　convergence　of　SA-AGA.　Experimental　results　of　four　test　functions　show　that　SA-AGA　has　superior　global　search　capabilities　and　higher　convergence　accuracy　than　other　comparative　algorithms,　and　can　effectively　avoid　premature　convergence.　©　COPYRIGHT　SPIE.　Downloading　of　the　abstract　is　permitted　for　personal　use　only.