基於感測資料曲線追蹤之雙足機器人斜面補償步態控制設計

Abstract

本論文之目的在於設計一雙足機器人於斜坡步行時的自主補償平衡系統，使機器人得以在未知斜坡上自行改變步態以免傾倒。機器人藉由身上安裝的感測器獲得環境資訊，並以達到平地步行時的感測器曲線為目的。藉由補償器的步態修正進行對感測資料的追蹤，使機器人在斜坡上感測器量測值能如平地所得，由此達到平衡行走的目的。系統架構方面，本論文中使用離線的步態設計以減低運算量，藉由比較步行時量測得到的資訊與平地時量測得到的資訊，以比較結果之誤差值，使用PID控制方式設計補償器的補償值，將補償值直接加入預先設計的步態以進行修正，藉此使機器人適應不同斜度的斜坡，進而順利步行。文中以NAO雙足機器人進行實驗，驗證所提出方法的可行性。經過不同斜度與不同步態實驗驗證，此系統可達成二足機器人於未知斜面上的穩定行走。

In human environments, biped robots are expected to walk on various types of ground surfaces. Balance control during bipedal walking on uneven terrain is a challenging problem. This thesis proposes a novel gait compensating control method that uses sensory data tracking to allow a robot to keep balance and walk on sloping surfaces. The real-time compensation control system adjusts the gait and makes online sensory data to track stored target sensory data, which are obtained from the robot when it walks on a flat surface. When the robot adjusts its gait and makes online sensory data similar to that for a flat surface, it achieves a stable walking pose. The proposed method is tested for several slopes and different walking trajectories, using the NAO robot. The robot automatically adjusts its pose on an unknown sloping surface, using the proposed method. Furthermore, experiments with transitional slopes further verify that the method can be extended to more general terrain variations.