In　this　paper,an　improved　artificial　potential　field(APF)method　combined　with　Bug2　is　proposed　for　dynamic　obstacle　avoidance　of　mobile　robots,which　eliminates　the　trajectory　oscillations　and　avoids　failure　when　encountering　a　complex　obstacle.First,the　problems　of　avoidance　failure　and　trajectory　oscillations　of　the　relative　position　and　velocity-based　APF　were　analyzed.Then,a　Bug2-based　APF　method　is　proposed　by　using　the　M-line　of　Bug2　to　optimize　the　avoidance　trajectory.The　proposed　method　makes　the　angle　of　the　virtual　repulsion　force　tangent　to　the　obstacle,which　results　in　avoidance　that　the　trajectory　surrounds　the　obstacle　along　with　the　M-line.Finally,the　effectiveness　and　practicability　of　the　proposed　method　are　demonstrated　by　performing　many　simulations　and　experiments.The　results　show　that　the　proposed　method　can　realize　safe　and　autonomous　dynamic　obstacle　avoidance　and　significantly　improve　the　robot＇s　success　rate　and　obstacle　avoidance　efficiency.