The　strategy　of　the　standard　Fruit　Fly　Optimization　Algorithm(FOA)　is　that　all　fruit　flies　search　locally　and　randomly　with　the　optimal　solution,　it　is　quite　easy　to　get　into　local　optimum;　Meanwhile,　as　the　step　size　is　fixed,　it　cannot　achieve　high　accuracy.　In　order　to　enhance　the　algorithm＇s　ability　to　avoid　falling　into　local　optimum　and　improve　the　accuracy,　this　paper　uses　different　tactics　to　deal　with　the　flight　strategy　and　step　size　of　fruit　flies　by　comparing　the　excellent　degree　of　candidate　solutions;　Furthermore,　by　adding　the　variance　threshold　of　concentration　value,　adopting　the　linear　generation　mechanism　of　candidate　solutions,　we　propose　a　Dynamic　Step　Size　Fruit　Fly　Optimization　Algorithm(DSS-FOA)　based　on　candidate　solutions.　Several　standard　test　functions　are　used　to　test,　results　of　the　test　confirm　that　DSS-FOA　is　superior　to　the　standard　FOA　in　global　optimization,　local　optimization,　accuracy　and　stability.　Besides,　the　algorithm　is　applied　to　the　PID　parameters＇　optimization　of　pH　in　wet　flue　gas　desulfurization(FGD)　system,　the　results　of　simulation　show　that　the　algorithm　has　excellent　effect　in　the　PID　parameters＇　optimization　of　pH　in　wet　FGD　system,　and　performs　much　better　than　the　standard　FOA.　©　2020　IEEE.