The　performance　of　a　network　can　be　enhanced　by　adding　links.　The　importance　of　the　rail　network　augmentation　scheme　comparison　is　brought　into　the　forefront　because　of　the　long　lead　time　and　disproportionate　high　cost.　This　work　focuses　on　the　influence　of　the　addition　link　among　the　initial　railway　stations　on　passenger　flow.　The　effectiveness　of　the　network　was　evaluated　by　flow　imbalance　(FIB),　average　transfer　times　(ATT),　average　path　length　(APL),　and　the　combined　evaluation　using　three　different　strategies,　namely,　passenger　flow　load　-　subgraphs　highest　node　(PFL-S),　passenger　flow　load　-　least　load　(PFL-L),　and　node　degree　and　betweenness　centrality　(Degree-BC).　The　case　study　on　the　real　urban　rail　network　with　2,211,468　trip　records　a　day　at　339　stations　show　that　PFL-S　and　PFL-L　have　better　performance　than　Degree-BC.　Simulation　experiments　reveal　that　the　link　construction　investment　at　the　early　stage　during　the　expanding　or　in　the　network　with　less　existing　transfer　stations　for　the　same　size　can　receive　greater　returns　than　those　nearly　completed　networks.　The　proposed　method　not　only　provides　a　policy　guidance　for　the　sustainable　development　of　rail　networks　but　also　offers　a　reference　for　other　networks　with　limited　flow　loading　capacity　and　high-cost　constraints.