Boarding　behavior　on　the　platform　of　high-speed　railway　(HSR)　station,　as　a　critical　problem　that　has　not　been　fully　explored　and　understood,　is　investigated　in　this　paper.　An　extended　cellular　automaton　(CA)　model　is　proposed　to　study　the　impacts　of　passengers’　heterogeneities　during　the　boarding　process.　Some　numerical　tests　are　conducted　to　explore　the　passengers’　movements　and　the　complex　traffic　phenomena　(e.g.,　passenger＇s　detour　behavior,　local　congestion　and　overtaking　behavior)　on　the　platform.　The　numerical　results　show　that　the　proportion　of　the　passengers　carrying　luggage　has　significant　impacts　on　the　boarding　process　(e.g.,　the　degree　of　congestion,　inflow　rate　at　the　entrance　and　route　choice).　Moreover,　the　heterogeneous　speeds　increase　the　total　boarding　time　and　decrease　the　boarding　efficiency　comparing　with　the　homogenous　speeds.　The　above　results　are　helpful　for　the　effective　management　of　pedestrian　flow　on　the　HSR　platform　and　the　optimal　design　of　HSR　platform　layout.　©　2020　Elsevier　B.V.