Autonomous　vehicles　will　bring　new　changes　to　the　road　traffic　flow.　The　driving　behavior　of　autonomous　vehicles　can　be　set　as　conservative　or　aggressive.　Therefore,　different　penetration　levels　of　such　autonomous　vehicles　in　traffic　can　lead　to　different　effects　in　road　traffic　flow,　which　is　also　directly　related　to　the　traffic　management　and　control　strategy.　Taking　typical　sections　of　freeway　as　research　objects,　this　paper　is　devoted　to　quantifying　the　traffic　impact　under　multiple　penetrations　of　autonomous　vehicles.　VISSIM　simulation　models　of　different　sections　of　freeway　are　built　for　different　traffic　status　scenarios　and　various　levels　of　autonomous　vehicle　penetration.　The　analysis　indicates　that　penetration　of　autonomous　vehicles　has　a　slight　positive　impact　on　traffic　operations　for　all　segments　in　the　free　flow　scenario.　In　the　saturation　flow　scenario,　parabolic　relationship　between　penetrations　of　autonomous　vehicles　and　average　travel　time　can　be　observed.　Moreover,　in　the　saturation　flow　and　wide　moving　jam　scenarios,　with　the　increase　of　penetration　of　autonomous　vehicles,　the　average　travel　time　decreases　in　merging　segment　and　waving　segment.　©　ASCE.