本文接續前人完成的輪腳複合式移動平台Quattroped,針對現有的機電架構上的缺陷做修正。並承接該平台的設計理念,開發出不同的運動模式。開發設計該機器人輪模式的演算法實現並驗證該平台在平地上的高效率移動能力;腳模式的演算法已準靜態的條件下做軌跡規劃及四足協調運動。利用四足機器人半圓腳機構,並透過幾何關係開發讓機器人在遭遇崎嶇地形時例如:自然崎嶇地、一階平台、橫桿甚至樓梯,能夠有效發揮腳的特性克服輪模式無法比擬的地形展現障礙能力。最後本文亦利用簡易的感測器,回授環境的訊息提供機器人步態的切換達到初步回授控制的效果;利用簡單的資訊以較低的成本達成自動化及智慧化的行為能力,進而提高實驗的成功率以及實用價值。
This paper continues the previous research of leg-wheel hybrid robot, Quattroped. We improve and modify the mechanical and electrical defects of the robot and carry on the design of the platform to develop different move locomotion. The wheel mode is constructed to let the robot move on flat ground efficiently. And The algorithm of trajectory planning and four leg coordination for quasi-static stair climbing in a quadruped robot is also reported. The development is based on the geometrical interactions by using the half-wheel legs of robot to overcome the rough terrains including step and bar crossing, irregular terrain passing, and stair climbing and to reveal the robot in legged mode has better mobility than wheel mode to cross obstacles or rough terrain. Finally, a basic feedback control on leg mode by simple sensor is also presented in this paper. Using simple information and low cost in order to achieve automation and intelligence capacity and to increase the success rate of experiments and practical value.