Gas　diffusion　layer　and　catalyst　layer　are　important　parts　in　a　proton　exchange　membrane　fuel　cell,　and　the　structure　parameters　can　affect　cell　performance.　Because　the　electrochemical　reaction　inside　the　cell　is　not　uniform,　it　is　particularity　important　to　explore　the　uneven　distribution　of　each　structure　parameter.　Thus,　a　fuel　cell　model　including　agglomerate　and　optimization　algorithm　is　built　to　study　the　optimal　step-wise　parameter　distribution.　These　parameters　include　cathode　gas　diffusion　layer　porosity,　platinum　loading,　ionomer　content　and　carbon　loading　inside　cathode　catalyst　layer.　The　four　parameters　are　calculated　comprehensively　to　obtain　the　optimal　three　segments　distribution　along　three　directions　of　the　coordinate　axis.　The　results　show　that　the　parameters　obtained　are　various　at　different　voltages　and　directions.　The　results　along　Z-Axis　has　the　best　positive　effect　on　cell　performance.　In　the　results　no　matter　which　direction　is,　the　maximum　peak　power　density　all　occurs　when　the　optimization　voltage　is　0.4　V.　The　cell　can　operate　in　wide　voltage　range　by　selecting　the　parameters　obtained　at　0.4　V　along　Z-Axis,　and　the　current　density　can　be　improved　at　least　1.859　%.　Bidirectional　mass　transfer　at　the　interface　between　two　domains　can　be　enhanced　for　the　model　with　optimal　parameters.