本論文主要探討蛇型機器人的行動策略，以現實中蛇的"側行式(sidewinding)"行動模式為原型，探討其行動原理並分析該運動模式的數學模型作為行動策略的基礎，再以 MIT 於2018年前後發表的" Cheetah 3 "機器人中不依賴視覺的"接觸檢測演算法(contact detection algorithm)"為啟發，透過電流感測器檢測流經馬達的電流大小，以卡爾曼濾波器推估其所受扭矩，達成接觸感測的狀態回授，以求進一步改善其行動效率以及對周遭環境的適應。;This thesis mainly discusses the action strategy of the snake-like robot. Taking the "sidewinding" action of the snake in reality as the prototype, it discusses its action principle and analyzes the mathematical model of the movement mode as the basis of the action strategy. Inspired by the "contact detection algorithm" on the "Cheetah 3" that does not rely on vision, the current flowing through the motor is detected by the current sensor, and the Kalman filter is used to estimate the torque it receives to achieve the state feedback of contact sensing, in order to further improve its operational efficiency and adaptation to the surrounding environment.
