수중 집광 로봇용 커버리지 경로계획 및 추적 제어기 설계

Abstract

This thesis proposes a coverage path planning method for an underwater mining robot to cover a given underwater region effectively and a tracking controller design to track the planned path generated by the coverage path planning method. The coverage path planning method includes preprocessing of a given 3D map, cell decomposition and minimal sum of altitude algorithm for minimizing turning number, an algorithm choosing coverage sequence of each cell, generating moving path between cells by modified sweeping line strategy to reduce an overlapped paths and produce a smooth path. For this work, the following tasks are implemented. Firstly, in preprocessing, a given 3D map is transformed to a binary map. Then, Morse decomposition method is used for decomposing the transformed binary map and minimal sum of altitude algorithm is applied to minimize a turning number. To define the coverage sequence of each cell, comparison between depth-first-searching algorithm and breath-first-searching algorithm based on spanning tree is proposed. After defining the coverage sequence, an entire coverage path and a moving path between cells using visibility graph algorithm are generated. For minimizing overlapped paths, modified sweeping line strategy is used when the length of overlapped paths is too long. Turning paths of the entire coverage path are smoothened. Secondly, an automated guided vehicle(AGV) with two passive wheels in front and back sides and two active driving wheels in left and right sides of the mobile robot is proposed. Kinematic modeling of this AGV system is presented. Based on the kinematic modeling, a backstepping controller is proposed to track the generated coverage path obtained from the proposed coverage path planning algorithm. Thirdly, to implement the proposed controller, a control system is developed based on microcontroller AVR ATmega128. A laser navigation system is used to measure the rotation angle and its position information of the proposed AGV. Finally, simulation and experiment results are presented to show the effectiveness and the applicability of the proposed coverage path planning algorithm and the proposed tracking controller for AGV system.