Development of an Android-based Autonomous Humanoid Robot Visual Obstacle Avoidance APP for FIRA HuroCup Games

• Main Authors: Szu-Yu Lin, 林思羽
• Other Authors: Chung-Hsien Kuo
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/59088364637962413370

碩士 === 國立臺灣科技大學 === 電機工程系 === 103 === This thesis presents an Android-based autonomous humanoid robot visual obstacle avoidance APP for FIRA HuroCup games. The proposed system is composed of a 7" pad computer and a microcontroller-based bipedal locomotion controller. The pad computer is responsible for collecting images, tracking a specific image target, processing fuzzy-logic based visual obstacle avoidance approaches and transmitting blue tooth commands for pan-tilt head control and bipedal locomotion commands. The microcontroller-based bipedal locomotion controller is responsible for controlling all joint motors to achieve desired bipedal locomotion commands in terms of the linear inverted pendulum model (LIPM).This study also employs the technique of JavaScript object notation (JSON) so that the locomotion parameters specified in the microcontroller-based bipedal locomotion controller can be easily desirable via a conventional web in a pad computer or a smart phone. Therefore, this thesis proposes an embedded bipedal humanoid robot control system without any uses of single board computer. This solution is compact and portable for most of conventional Android-based pad computer or smart phones. Finally, obstacle avoidance games specified in the FIRA HuroCup league were used to demonstrate the effectiveness of the proposed embedded autonomous bipedal humanoid robot control architecture.