透過您的圖書館登入
IP:3.149.26.176
  • 期刊
  • OpenAccess

應用于農業機器人的自走式車輛的研制

Development of an Autonomous Vehicle for Agricultural Robotics

摘要


自走式車輛可以作為移動式機器人的載具或是無人搬運車,執行農業自動化的生產和運輸的工作。本文說明一具可應用於農業機器人的自走式車輛的硬體架構與其導航控制的方法和性能。此車輛由機體、影像系統、及控制系統所組成。機體長71.5cm,寬43.5cm,最大行走速度為0.25m/s。影像系統的解析度是64*256 點。影像資料利用記憶體直接取存技術收集並儲存。控制系統具備以影像資料定位及導航的功能,並預留輸出入介面以連接其他感測器或驅動裝置。車輛的導航是將預設路徑以折線表示,經由定位系統得到目前車輛的位置及方向後,以循跡歸屬確認車輛所屬線段的管制區,再經由偏移量分析得到位置及方向偏差量,以之計算所需的車輛速度及差速信號,控制車輛的行進。循跡性試驗的結果,本車輛以10cm/s的速度行駛直線時,橫向偏移在±0.5cm以內,在20cm的起始偏移下,行駛直線距離60cm內即可修正至1cm內。行走速度提高時修正偏移所需的行走距離亦增加。作直角轉彎時,軌跡在直角端點前後約12cm以內會直接穿越而不經過直角頂點。重覆性測試的結果,在10cm/s的直線行駛速度下,停止點平均距離誤差為0.9cm,而車輛軸線方向的誤差大於橫向偏移。

並列摘要


Autonomous vehicles may be used as mobile robot carriers or automatic conveyors for the automation of agricultural production and transportation. This paper describes the development of an autonomous vehicle for agricultural robotics. The vehicle is composed of the chassis, an image system, and the control system. The chassis is 71.5 cm long and 43.5 cm wide with a maximum travel speed of 0.25 m/s. The image system has a resolution of 64*256 and can digitize images and store them onto the mother board using direct memory access. The control system could guide the vehicle using the image information and connect to other sensors or actuators. The vehicle path is pre-specified by a sequence of straight lines. During the navigation control, the vehicle analyses its position and heading angle, locates itself on the path, calculates the position and heading deviations, and follows its path by sending velocity and differential signals to control its movement. Tracking results show that the vehicle could follow straight lines with deviations less than ±0.5 cm at a speed of 10 cm/s. From an initial offset of 20 cm, it could move back to the preset course to within 1 cm in 60 cm of straight travel distance. The travel distance for correcting the initial offset increases when vehicle velocity increases. During right angle turns, the vehicle would take a short cut across from about 12 cm before and after the turning point. Repeatability tests show that the average error of stop points was about 0.9 cm, while the errors in the axial direction were larger than the side deviations.

並列關鍵字

Autonomous vehicle Image system Guidance Robotics

延伸閱讀