In　order　to　reduce　the　energy　consumption　of　train　operation,　an　optimization　method　based　on　genetic　algorithm　of　golden　section　is　proposed.　Firstly,　the　Multi-particle　train　model　is　established.　Secondly,　the　optimal　operation　strategy　of　subway　trains　is　analyzed　according　to　different　ramps.　Then,　a　golden　section　genetic　algorithm　(GR-GA)　is　proposed　to　solve　the　problem　that　genetic　algorithm　is　easy　to　fall　into　local　optimum.　A　golden　section　genetic　algorithm　(GR-GA)　is　proposed　to　search　for　the　optimal　transfer　position　of　train　and　the　best　adaptive　point　of　searching　crossover　and　mutation　operator　with　golden　ratio　is　introduced,　which　improves　the　local　optimization　ability　and　convergence　performance.　Taking　Yizhuang　line　as　a　simulation　case,　the　results　show　that　the　proposed　algorithm　has　a　better　optimization　effect.