Weaving　area　is　one　of　the　typical　traffic　bottlenecks　of　urban　expressway　and　its　traffic　dynamics　are　very　complex.　With　the　development　of　vehicle　intelligence　and　network　technology,　automatic　vehicle　driving　and　real-time　dynamic　communication　become　possible,　which　provides　technical　support　for　the　control　of　vehicle　group　behavior　in　weaving　areas.　In　this　paper,　the　cooperative　lane　changing　among　on-ramp　vehicles,　off-ramp　vehicles　and　straight　driving　vehicles　in　weaving　areas　of　the　urban　expressway　is　studied.　Based　on　the　analysis　of　traffic　characteristics　in　weaving　areas　under　the　environment　of　connected　and　automated　vehicles　(CAVs),　four　cooperative　strategies　for　vehicle　groups　in　the　weaving　area　based　on　different　road　spatial　flow-balance　are　proposed.　The　simulation　model　of　the　weaving　area　is　built　with　VISSIM　simulation　software,　and　the　proposed　optimization　schemes　are　simulated　and　compared.　The　results　show　that　the　four　strategies　have　similar　effects　when　the　traffic　volume　on　the　mainline　and　on-ramp　are　both　low.　In　other　states,　the　strategy　of　off-ramp　vehicles　preferentially　distributed　in　the　outer　lane　is　superior　to　the　strategy　of　off-ramp　vehicles　with　different　distribution　ratios　in　each　lane　of　the　mainline,　that　is,　the　strategy　of　inducing　or　controlling　all　off-ramp　vehicles　to　the　outermost　lane　in　advance　before　entering　the　deceleration　lane　(for　example,　the　ratio　of　off-ramp　vehicles　in　three　lanes　is　0:0:1).　This　paper　intends　to　provide　theoretical　support　for　reducing　vehicle　conflicts　and　improving　traffic　safety　and　traffic　efficiency　in　urban　expressway　weaving　areas　under　the　CAV　environment.　©　2020　SAE　International.