仿生六足機器人越障步態與跳躍步態開發

Abstract

本文介紹RHex型式的六足機器人整體架構、感測器及機電系統，並分析不同步態的行走狀況及利用此步態調整達轉彎效果，此論文可分為兩大部分，第一部分是以蟑螂變換步態跨越較高障礙為靈感，設計執行全自動越障操作，讓機器人可靠地爬越230%腳長高度的障礙，兩個紅外線距離感測器和傾斜儀的使用來偵測障礙出現、障礙與機器人位置關係、計算障礙高度和辨認障礙種類，讓此機器人自動調態。第二部分是開發跳躍行為讓機器人跳躍過狹窄障礙和溝渠，經由簡化腳模型建立三組腳的跳躍模型分析跳躍表現並找出理想跳躍方式，更進一步加裝紅外線偵測障礙距離自動變換步距達適當位置做跳躍動作。最後經由實驗評估模型和演算法的表現及可行性。

This thesis introduces the composition, sensor, mechatronics system for a RHex-style hexapod robot, and then analyzes walking condition in different gait design, moreover, it uses this gait adjustment to achieve the goal of turning. Besides, This thesis develops two kinds of gait to fulfill particular missions. Firstly, inspired by the observation that the cockroach changes from a tripod gait to a different gait for climbing high steps, this thesis reports on the design and implementation of a novel, fully autonomous step-climbing maneuver, which enables robot to reliably climb a step up to 230% higher than the length of its leg. Two infrared range sensors and an inclinometer are installed on robot to detect the presence of obstacle, its orientation relative to the robot’s heading, compute obstacle height and recognize different kinds of obstacle, thus enabling the robot to adjust its gait automatically, in real time. Secondly, a jumping gait development makes robot jump over a narrow obstacle and a ditch. A jumping model is established by a simplified legged model to analysis jumping performance to find the ideal jumping method, and further, it uses infrared range sensor to detect the distance between robot and obstacle then changes its gait to achieve ideal position accomplishing jumping performance. Finally, the behavior of the algorithm and model are evaluated experimentally.