基於量子基因演算法之自主導航系統設計

本論文透過量子基因演算法(Quantum Genetic Algorithm，QGA)設計一套機器人之自主導航系統。此自主導航系統利用QGA產生多樣性的量子物種來模擬自然界物競天擇的特性，經過多次迭代計算後可以得出最佳路徑，之後再結合機器人操作系統(Robot Operating System， ROS)實現自主導航系統。本論文使用了機器人作業系統(Robot Operating System，ROS)所提供開源的Navigation Stack套件，來實現導航系統。Navigation Stack套件包含定位、感測器以及路徑規劃等功能，而且有設計API程式去對應ROS系統。在研究方法部份，本論文分成(1)單一目標之路徑規劃、(2)基於量子基因演算法路徑最佳化設計和(3)導航系統整合等三個項目進行研究。在路徑規劃上面，本論文透過全域地圖和區域地圖的搭配，可以提高對於地圖環境變動的容忍度，再使用ROS所提供的開源套件實現機器人自主導航。在導航系統部份，本論文設計一執行策略介面去做全面性的整合，讓商店機器人可以透過操作介面接收定位點之後，自主導航到目標且執行動作，最後實現在實驗室自製移動平台上面。

關鍵字

移動平台；導航系統； Gazebo模擬器；量子基因演算法；路徑規劃；最佳化問題；旅行銷售員問題

並列摘要

This paper designs an autonomous navigation system for a robot through the Quantum Genetic Algorithm (QGA). This autonomous navigation system uses QGA to generate a variety of quantum species to simulate the characteristics of natural selection. After many iterative calculations, the optimal path can be obtained, and then combined with the Robot Operating System (ROS) to achieve autonomy Navigation System. This paper uses the open source Navigation Stack suite provided by ROS to implement the navigation system. The Navigation Stack kit includes functions such as positioning, sensors, and path planning, and has designed API programs to correspond to the ROS system. In the research method part, this thesis is divided into three projects: (1) single target path planning, (2) path optimization design based on quantum genetic algorithm and (3) navigation system integration. In terms of path planning, this paper can improve the tolerance of map environment changes through the combination of global map and regional map, and then use the open source kit provided by ROS to realize autonomous navigation of the robot. In the part of the navigation system, this paper designs an implementation strategy interface for comprehensive integration, so that the robot can automatically navigate to the target and perform actions after receiving the positioning point through the operation interface.