Scholarlink

<p>To address the control problem of multi-legged robots for future extraterrestrial exploration, a hierarchical control framework based on model predictive control and virtual model control (MPC-VMC) is proposed. First, the upper-level trajectory tracking controller establishes the discretized error state space equations for the planar motion of the quadruped robot. An MPC-based controller is designed to consider speed and acceleration constraints and quickly solve for the optimized body’s center of mass (CoM) speed and yaw angle. These values are then passed to the middle-level controller as desired inputs. Then, in the middle level, the dynamic relationship between the body’s motion state and the ground reaction forces is described. Utilizing VMC principles, a motion controller is designed to calculate the ground reaction forces. Simultaneously, the gait generator produces gait timing sequences and desired foot-end trajectories. Next, the leg dynamics model is established. The lower-level leg controller is designed based on VMC, incorporating dynamic feedforward and compliant control to enhance terrain adaptability. Finally, the proposed hierarchical control framework is verified through simulations and experiments. The results show that the framework is able to avoid obstacles and adapt to irregular terrain, balancing control accuracy and computational efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hierarchical control framework for multi-legged extraterrestrial exploration robot based on model predictive and virtual model control

  • 期刊论文

  • 发表日期：

    2026年4月11日

查看全文
  • Qingsheng Wei,
  • Shigang Peng,
  • Xiang Cheng,
  • Cheng Wei,
  • Jian Tian

摘要

To address the control problem of multi-legged robots for future extraterrestrial exploration, a hierarchical control framework based on model predictive control and virtual model control (MPC-VMC) is proposed. First, the upper-level trajectory tracking controller establishes the discretized error state space equations for the planar motion of the quadruped robot. An MPC-based controller is designed to consider speed and acceleration constraints and quickly solve for the optimized body’s center of mass (CoM) speed and yaw angle. These values are then passed to the middle-level controller as desired inputs. Then, in the middle level, the dynamic relationship between the body’s motion state and the ground reaction forces is described. Utilizing VMC principles, a motion controller is designed to calculate the ground reaction forces. Simultaneously, the gait generator produces gait timing sequences and desired foot-end trajectories. Next, the leg dynamics model is established. The lower-level leg controller is designed based on VMC, incorporating dynamic feedforward and compliant control to enhance terrain adaptability. Finally, the proposed hierarchical control framework is verified through simulations and experiments. The results show that the framework is able to avoid obstacles and adapt to irregular terrain, balancing control accuracy and computational efficiency.