![]() | Only 14 pages are availabe for public view |
Abstract Humanoid robots will have many applications in the industrial sector in near future. Based on this motivation, the thesis presents a complete study on humanoid robot kinematics and dynamics. The thesis also discusses some basic methods for walking pattern generation and elaborates on some modifications to enhance the basic walking pattern generation methods in terms of stability and power consumption optimization. The thesis then proceeds to introduce some stabilization techniques using ZMP feedback and upper body posture correction. The suggested algorithms were validated using MATLAB [104] simulations for NAO robot. The presented methods produced satisfactory results in terms of stability, as the RMSE for ZMP x-coordinates (x-axis being parallel to walking direction), ZMP y-coordinates, upper body roll angle deviation and upper body pitch angle deviation were found to be 0.0028 m, 0.0139 m, 0.0028 rad and 0.0014 rad respectively. Also, the presented work contributed significantly to humanoid walking power consumption optimization as the average locomotion cost of transport was reduced from 0.2367 to 0.11993. |