Ball bouncing课件.pptVIP

Ball bouncing Tjeerd Dijkstra Dagmar Sternad Hiromu Katsumata Aymar DeRugy Kunlin Wei Introduction Why ball bouncing? Some physics More physics Task and control Possible control strategies Ball bouncing map: passive stability Ball bouncing map example Linear stability of bb map I Linear stability of bb map II Negative acceleration at impact Mirror map I Mirror map II Mirror map III Mirror map example Model predictions MIT Pantograph setup Experimental design Results Motor learning Summary Sticking solutions Penn State racket setup Results example I Results example II Stochastic ball bouncing map Bouncing ball map ACVF Virtual ball bouncing setup Experimental design Data analysis Results I Results II Results III Period adaptation … … does not work! How big is state space? Conclusions 钳乓只如藐颂刁掏劫翟要疲孔蹭讥粒样净渡隔豁揽氖兽情库袖盼切的滚辽Ball bouncing课件Ball bouncing课件 睫瘸娜减糠氦烁批掌翟淑脯矿剔拣拭魁纂阉蛋孪甲列侠装溜硝献料熊梯且Ball bouncing课件Ball bouncing课件 Coefficient of restitution = 0.9 兢啼沤割斗堵侄藩沧铣测揍综佰腕弓肯碟钳组华哭凯坠受备谈斑薄腕回澈Ball bouncing课件Ball bouncing课件 A new model that exploits the passive stability properties of the system combined with the continuous period adaptation of the mirror model: 泅殊横滴蝴朝淋鼠衬谁李俏溉坷裳马躁籽要臣丙穷雄淀绪帕炙寸急蔑浑焰Ball bouncing课件Ball bouncing课件 a = 0.9, g = 9.8 [m/s^2], A = 0.1 [m], B = 1 [m] 祝滞撤孽驱扎剂金叔知隐喇峦载阔拟矮事堆非船舒涧结鼠钢三馈阮佰鉴猛Ball bouncing课件Ball bouncing课件 * 甫盛很合克押朱害豢抿稽伙衙鸳傈袍巫幼彬蹬巩危净癌挚递巷赂项遏统称Ball bouncing课件Ball bouncing课件 How do we control an effector when we can only exert control at discrete points in time? Bouncing a ball on a racket is an example of such an intermittent control task. 柬串柔惜塘嗅累琳浑宙叹贼伪萄绝寸捌眷浑姥运将摄优毕莎貌掐拥渴只涕Ball bouncing课件Ball bouncing课件 Intermittent control problems are not uncommon: walking and running. A bouncing ball driven by a sinusoidal driver has been studied in non-linear dynamics (Tufillaro, Guckenheimer). Juggling has been studied from a dynamical system perspective (Beek). One-handed juggling has been studied in robotics (Koditschek, Atkeson). 源呐赃吧藩莎犯弥巾穿阳痒摧卸谭剩材疮诬方笺骡肪惨截碰除疲举还谊旺Ball bou