Mixed-traffic　crosswalks　with　disordered　e-bikes　are　one　of　the　most　common　conflict　zones　in　urban　traffic.　The　frequent　interactions　between　pedestrians　and　e-bikes　may　result　in　a　reduction　in　traffic　efficiency　and　safety　at　intersections.　It　is　necessary　to　understand　nonlinear　dynamics　of　the　mixed　traffic　flow.　This　study　presents　a　modified　social　force　model　to　describe　complex　phenomena　of　pedestrian-e-bike　mixed　flow　at　a　signalized　crosswalk.　Geometry　of　e-bikes,　visual　range　and　typical　behaviors　(i.e.,　avoidance　behaviors)　are　introduced　in　our　model,　and　compared　with　an　observational　experiment.　The　proportion　of　e-bikes　and　the　impact　of　the　lane　of　e-bikes　on　pedestrians＇　walking　comfort　and　safety　are　discussed.　The　arrival　rate　of　e-bikes　and　the　impact　of　the　setting　of　the　lane　of　e-bikes　are　analyzed.　It　is　proved　that　pedestrians＇　comfort　decreases　with　the　increasing　proportion　of　e-bikes.　The　pressure　of　the　pedestrian-e-bike　mixed　flow　is　higher　than　that　at　a　conventional　crosswalk　without　e-bikes.　The　maximum　crossing　time　of　e-bikes　decreases　with　the　increasing　width　of　the　lane　of　e-bikes.　Correspondingly,　when　the　width　of　the　lane　of　e-bikes　increases,　the　average　speed　of　e-bikes　increases.　Results　suggest　that　an　appropriate　width　of　the　lane　of　e-bikes　has　a　positive　influence　on　mixed　flow.　The　study　is　helpful　to　an　in-depth　understanding　of　nonlinear　dynamics　in　pedestrian-e-bike　mixed　flow,　and　is　beneficial　to　safe　pedestrian　facility　design.