Connected　and　automated　vehicles　(CAVs)　and　human-driven　vehicles　(HDVs)　are　expected　to　form　the　mixed　traffic　flow　in　recent　decades.　This　paper　studies　the　control　problem　of　the　mixed　traffic　flow　at　a　signalized　intersection.　A　two-level　cooperative　control　method　is　proposed　to　optimize　both　intersection　signal　timing　and　trajectories　of　CAVs.　At　the　upper　level,　to　minimize　total　travel　delay,　dynamic　programming　is　used　to　find　the　optimal　phase　scheme　at　the　beginning　of　each　signal　cycle　and　assign　arrival　time　to　vehicles.　At　the　lower　level,　CAVs　trajectories　are　generated　from　downstream　to　upstream　of　the　control　zone　in　sequence　based　on　optimal　control　with　the　objective　of　driving　comfortably　and　economically.　Numerical　experiments　are　conducted　in　the　traffic　simulation　platform　SUMO,　showing　that　the　proposed　method　can　reduce　travel　time　and　fuel　consumption　effectively　compared　with　the　traditional　actuated　control.　©　2023　Technical　Committee　on　Control　Theory,　Chinese　Association　of　Automation.