当前位置: X-MOL 学术Adv. Theory Simul. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Jumping Motion Based on Momentum Optimization Control and Differential Dynamic Programming for a Multilocomotion Robot
Advanced Theory and Simulations ( IF 3.3 ) Pub Date : 2022-02-24 , DOI: 10.1002/adts.202100636
Guoshuai Liu 1 , Zhiguo Lu 1 , Ruchao Wang 1 , Haibin Zhao 1 , Chong Liu 1 , Xinhao Shi 1
Affiliation  

Jumping motion is an indispensable motion mode for a multilocomotion robot, which will improve the crossing ability of the robot. This paper proposes a novel jumping control scheme for an MLR to perform jumping motion. The control strategy divides the whole jumping motion into three phases: the launch phase, the flight phase, and the landing phase. In the launch and landing phases, the controller is designed based on the centroid dynamics. By controlling the centroid momentum, the robot can achieve the desired centroid trajectory to obtain the initial take-off speed in the launch phase and to maintain stability and reduce impact in the landing phase. In the flight phase, Differential Dynamic Programming algorithm is used to adjust the posture of the robot to further improve the jumping height and prepare for a smooth landing. Finally, the effectiveness of the proposed method is evaluated through simulations.

中文翻译:

基于动量优化控制和差分动态规划的多运动机器人跳跃运动

跳跃运动是多运动机器人不可缺少的运动方式,它将提高机器人的穿越能力。本文提出了一种新的跳跃控制方案,用于 MLR 执行跳跃运动。控制策略将整个跳跃运动分为三个阶段:发射阶段、飞行阶段和着陆阶段。在发射和着陆阶段,控制器是基于质心动力学设计的。通过控制质心动量,机器人可以在发射阶段获得所需的质心轨迹,从而获得初始起飞速度,并在着陆阶段保持稳定并减少冲击。在飞行阶段,采用差分动态规划算法调整机器人姿态,进一步提高跳跃高度,为平稳着陆做准备。最后,
更新日期:2022-02-24
down
wechat
bug