System Design and Implementation of a Home-Service Robot

• Main Authors: Chia-Ming Chen, 陳家銘
• Other Authors: 蔡清池
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/30206514487588900688

碩士 === 中興大學 === 電機工程學系所 === 95 === This thesis presents methodologies and techniques for self-localization, safe patrolling, motion planning and system-on-programmable-chip (SOPC) implementation of a home-service robot with a manipulator equipped on the omni-directional mobile platform. A self-localization algorithm based on the triangular approach is proposed using the omni-directional vision system. Safe patrolling in house environment is achieved by two kinds of path planning, including global path planning and local path planning. The global optimal path is determined using a novel encoded genetic algorithm which provides the home-service robot with a static obstacle-free path. With the global optimal path, a weight decision method based on the laser measurement system is presented to steer the robot follow the local path with obstacle avoidance. Being a home-service robot, it is designed to have two functions, object fetch-and-carry, and fire extinguishment. The 3-D position of the fire is estimated by binocular webcams, and a proposed elite genetic algorithm is developed to find the optimal configuration of the mobile manipulator moving from one point to another when executing the fire extinguishment. A small-scale personal computer is employed to deal with environment estimation from the laser scanner, and image processing from the omni-directional vision system and the binocular vision system. Both kinematic control of the manipulator and the omni-directional mobile platform are respectively performed by a Stratix II Nios development board, and a Stratix Nios development board. Numerous simulations and experimental results are conducted to verify the effectiveness and merits of the proposed methods.