應用模糊控制於平行停車作業

Abstract

本研究主要設計並製作雙輪致動型的車型機器人，藉由低成本的感測元件輔助控制器來完成平行停車之任務實驗，此車型機器人是一具完全獨立的個體其中包含感測元件、微控制器與伺服馬達致動器。在平行停車過程中，隨時透過感測器之訊號來判斷車體之相對位置，對應所設計之規則庫來判斷控制指令直到完成任務。此自走車之感測器包含紅外線感測器與電子羅盤，所採用之控制器是採用帕勒斯公司(Parallax, Inc.)所開發之微控制器(BASIC Stamp 2)，驅動馬達採用連續轉動型之伺服馬達。本研究應用模糊控制理論為基礎，設計一個以模糊邏輯為核心的控制器，在進行硬體實驗前先經由套裝軟體(Matlab)模擬測試以證明此方法之可行度。硬體實驗結果顯示同一控制方法可讓自走車完成停車於不同空間，其最小需求長度為車體長度之1.15倍。

The purpose of this study is to design and make a two-wheel actuated car-like mobile robot (CLMR). In order to save money, low cost sensing components are used and assisted by a micro-controller to accomplish the parallel parking task. CLMR is an autonomous robot containing sensor components, micro controller and servo motor actuator. In the process of parallel parking, the sensor signals constantly determine the relative position of the body, and then correspond to the designed rule base to decide the controlling instructions until the completion of the task. The autonomous robot consists of the following four subsystems which are two infrared sensors, an electronic compass, a central processor unit, and a driving device. The central processor unit (BASIC Stamp 2) is developed by Parallax, Inc. The driving device consists of servo motors which rotate continuously. Rule base of the controller is designed based on fuzzy logic. In order to prove its practicality, the package( Matlab) is used to simulate before the hardware testing. The results of hardware experiment indicate that CLMR accomplishes the parking task in different spaces with the same method. The minimum length requirement is 1.15 times to its body length.