The　optimization　of　passenger　evacuation　path　in　subway　station　under　emergencies　directly　affects　the　evacuation　efficiency.　Aiming　at　the　evacuation　scenario　of　passengers　in　subway　station　under　multiple　fires,　a　path　planning　method　based　on　multi-objective　robust　optimization　is　established　in　this　paper.　The　shortest　total　evacuation　time,　the　minimum　total　risk　and　the　minimum　total　congestion　cost　are　taken　as　the　objectives　where　the　robustness　of　travel　time　and　risk　is　also　considered.　NSGA-II　algorithm　is　used　to　solve　the　model,　the　optimal　Pareto　solution　under　a　certain　robust　control　parameter　is　obtained　according　to　the　principle　of　minimizing　the　total　cost　function,　and　the　overall　optimization　degree　of　route　is　obtained　by　using　the　evaluation　index.　Taking　Qingdao　May　4th　Square　subway　station　of　China　as　a　verification　example,　a　fire　diffusion　model　is　built　by　using　Pyrosim,　and　a　passenger　evacuation　system　is　built　by　using　Pathfinder.　The　multi-objective　robust　path　optimization　model　of　passenger　evacuation　is　applied　to　the　built　passenger　evacuation　system　under　fires.　The　simulation　results　show　that　the　overall　optimization　degree　of　route　can　reach　12.8%　,　when　the　time　and　risk　robust　control　parameters　are　30　and　30,　respectively.　The　method　can　be　used　to　guide　the　passenger　evacuation　in　subway　stations　under　multiple　hazard　sources　and　improve　the　safety　index　of　passenger　travel.