With　the　development　of　sensor　and　network　technologies,　a　type　of　hybrid　sensor　network　(HSN)　has　been　developed　in　traffic　information　acquisition　systems.　Sensor　types　in　HSNs　are　diversified,　mostly　by　batteries　or　solar　energy　as　their　energy　source.　Energy　determines　their　lifetime.　If　energy　consumption　of　sensors　can　be　reduced　from　the　aspect　of　communication,　their　energy　consumption　can　be　saved　and　lifetime　can　be　increased,　so　as　to　improve　the　benefit　and　efficiency　of　the　whole　network.　To　optimize　the　efficiency　and　consumption　of　HSNs,　a　graph　theory-based　model　was　proposed　by　considering　nodes＇　connectivity　power　and　the　energy　consumption　of　data　transmission.　Targeting　different　data　volumes,　two　undigraph-based　algorithms　(PMST　and　DSPA)　were　introduced　to　solve　the　proposed　optimization　model.　For　a　low　data　volume,　the　minimum　spanning　tree　network　based　on　PMST　will　be　the　optimal　topology　of　an　HSN.　For　a　high　data　volume,　the　shortest　path　network　based　on　DSPA　will　be　the　optimal　topology.　A　numerical　example　of　comprehensive　traffic　information　acquisition　at　a　signalized　intersection　was　used　to　verify　the　proposed　algorithms.　The　solution　process　and　final　optimal　topologies　based　on　the　PMST　and　the　DSPA　were　shown.　Moreover,　the　energy　parameters　of　different　optimal　topologies　of　HSNs　were　discussed,　and　it　was　proved　that　the　optimal　topology　of　an　HSN　is　dependent　on　the　data　volume　of　each　sensor　node.