基於行為模式之家用機器人導航設計

Abstract

本論文提出一個以行為模式作為基礎的導航架構，以模糊邏輯控制器作為整個導航架構中主要的設計方法。文中提出一個稱為超音波方向權重的觀念，可以讓我們用非常簡單且非常少的模糊控制規則就能設計出我們的導航架構，因此系統計算速度相當快。我們利用一個啟發性網路使所設計的超音波方向權重值，可以隨著環境的改變而有自行調整的能力。本論文藉由電腦模擬與機器人導航實驗的結果，來驗證我們所設計的導航架構確實有執行導航的能力。

A home robot navigation system has been designed and implemented based on a behavior-based approach. In this thesis, we propose a concept of direction weight of robot environment sonar scanning. This concept facilitates the design of navigation system using fuzzy logic controller. Only a few fuzzy logic control rules are used in the navigation behavior design. The system has fast response in an unknown environment. Fuzzy–Kohonen Clustering Network (FKCN) is adopted for direction weight determination. In this method, the direction weight is adapted to the environment change. We will show both the computer simulation and experiment results to verify the navigation system has the capability of robot navigate.