運用雷射測距儀之機器人定位設計

Abstract

智慧型機器人必須有在移動環境中自我定位的能力，以達成各種實際應用。本論文的目的在於設計一機器人之自我定位系統來估測機器人在環境中的姿態。本論文利用粒子濾波器設計位置估測的方法，藉由里程計的資訊先對機器人的位置做預測，再經由從距離感測器對環境的量測結果得到的機器人位置資訊來對預測的機器人位置做修正，最後由粒子濾波器估測出機器人的位置。本論文並以將感測資訊與環境地圖做比對的方式估測朝向角，將雷射掃描儀量測的環境距離資料以最小平方法擬合成包含若干線段的線段組，再結合機器人的位置資訊與內建的環境地圖做比對，藉由比對結果可以估測出機器人的朝向角。本論文使用的距離感測器為雷射掃描儀，其具有相當高的精確度與可靠度，利用它來得到用來擬合成線段的環境距離資料。最後經由機器人繞行環境實驗的驗證，結果顯示所設計的機器人之定位系統確實有預期的功效。

Mobile robots should have the ability of self-localization for navigation in an environment. The objective of this thesis is to design a mobile robot localization system using a laser scanner. A particle filter is designed to estimate robot’s position in this study. The particle filter predicts robot’s position from odometer data, and updates the predicted robot’s position by using the distance measurements of the environment. The localization system estimates robot’s orientation by comparing the information from laser range finder with the environmental map. We can estimate the robot’s orientation by first fitting the distance data from laser range finder to a line set, then combining the line set with robot’s position information to match with the environment map. The system has been demonstrated using an experimental mobile robot. Several navigation experiments in an indoor environment have been carried out. Satisfactory experimental results are presented in this thesis.