Development of Vision-Servo Dynamic Positioning System for mobile robot

• Main Authors: Che-Kuei Hsu, 許哲魁
• Other Authors: Kuo-Ching Chiu
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/6k8nj2

碩士 === 國立虎尾科技大學 === 車輛工程系碩士班 === 105 === This paper proposed a novel control scheme for dynamic positioning of mobile robot. With an imaging-based surveying method, the positions of the mobile robot relative to the working space can be located by utilizing two vision sensors and some target poles. The robot can measure the distances to two fixed target poles by controlling the optical axes of two rotatable on-board vision sensors to aim at the target poles respectively, and acquires the angles of the intersections between the longitudinal axis of the robot and optical axes of the sensors. After extracting the object images, the 2-D geometric relation between the target objects and the robot can be described by coordinate transformation. Based on this geometric information and path planning, an adaptive neural sliding mode control (ANSMC) algorithm can be derived by the Lyapunov stability theory for the Microcontroller (Arduino-MEGA) to produce the control command for keeping the mobile robot on the planned path. For verifying the performance of the proposed strategy, a reciprocal path planning for mobile robot was carried out to achieve fully coverage of working space. The experiment results indicated that our method outperforms the methods based on a compass/GPS sensors. The maximum error between the estimated location and the real location is only about 5 cm per 10 meters, which meets the standard for engineering applications.