Motion Control And Gait Planning For Humanoid Robot

Humanoid robot is a complex multi-rigid-body biped walking system, which has multiple-input multiple-output, strong coupling, nonlinear and other characteristics. This not only requires its motion control system having high performance, but also requires high performance of gait planning algorithm using to achieve stable walking.The paper discusses the advantages and disadvantages of the present humanoid robot motion control system, according to the smaller, light-based, modular,open design and low cost requirements, a humanoid robot distributed motion control system proposed. The system uses hierarchical control architecture, mainly including the console, real-time computer control systems and the multi-joint controller which based on DSP. The console communicates with the real-time computer system through a wireless Ethernet, and the real-time computer system communicates with multi-joint controller through CAN bus. By applying the system to MHI-I, the experiment results not only showed an excellent control performance of the system, but also demonstrated the system has a series of advantages, such as: a small, modular, open structure, supporting network and so on. In order to prevent the impact of different body posture and joint initial position on the motion control, the paper presents the design and implementation of a humanoid robot home posture precise reset system after a deep research on the advantages and disadvantages of current humanoid robot home position reset system. Through experimenting on MIH-1, the system was verified not only be able to achieve a stable, fast and accurate home posture, but also could detect the loose joints, eliminate the effect of the backlash of joints and zero-position drift to the initial position.The paper studies the humanoid robot gait planning algorithm later. After the kinematics and inverse kinematics model of MIH-I have been established, the gait planning algorithm based on ZMP were achieved by Matlab. In order to protect the physical prototype of humanoid robot, simulating gait planning algorithm by Adams first and applying the algorithm on the physical prototype of humanoid robot later is proposed. And this method works effectively on protecting the real robot. At last, the reasonableness of the method is verified by applying the gait planning algorithm on both Adams and MIH-I correspondingly, and stable walking of MIH-I has been achieved.