本文之主要目的為增強人型機器人在行走時的穩定性。透過結合ADAMS/Control和MATLAB/Simulink的動態模擬器,進而模擬和分析人型機器人的行走狀態以及其穩定性。並類比於人類步行的模式,分析影響機器人行走的各個因素,做一整合性的探討。 在機器人步伐行走方面,首先藉由分析機器人行走在不同環境下的步伐參數,以及透過零矩點(ZMP)的軌跡規劃來達到平滑的行走模式。模擬的結果顯示,適當的步伐參數以及零矩點規劃能大幅降低能量的耗損並增加機器人在行走時的穩定性。 在機器人行走的即時控制上,藉由模仿人類的運動模式,進而發展關聯性控制(correlation-based control),透過所發展的控制方法可依據零矩點的回授訊號,即時改善人形機器人在行走時的穩定性。本文另發展一套RWLS(Robust Weighted Least-Squares)的演算法,以求得穩定的逆運動學(inverse kinematics)之解。最後透過動態模擬器的模擬結果加以驗證。
This thesis aims at the mobility enhancement of a humanoid robot. Through co-simulation method implemented by using ADAMS/Controls and MATLAB/Simulink, we analyze the kinematics and dynamics of a humanoid robot. Moreover, the analysis of the walking factors on different environment and the walking pattern generation analog to human locomotion will be thoroughly discussed. In this thesis, we developed a simple method to generate the foot trajectory even in different environment. And a brief investigation of foot parameters is made. In order to achieve smooth walking pattern generation, the Zero-Moment Point (ZMP) trajectory planning is proposed in this thesis. With simulation results, we can tell the energy consumption decreases and the robot walks more stably with planning ZMP trajectory and suitable foot parameters. We also develop a correlation-based control (CBC) to realize on-line COG trajectory planning. The correlation-based control is designed by the correlation of those factors which can inference the walking stability. Moreover, the RWLS (Robust Weighted Least-Squares) is also developed for the reliable inverse kinematics solution. Finally, the simulation results display that our algorithms can efficiently enhance the stability of the humanoid robot.