衛星導航應用於自走式農機轉向系統之研究

Abstract

在工業環境自動化的現代，農業機械也持續朝向自動化的方向進步著，本研究因而致力於開發半自動農業機械，目標為建立適合台灣農業機械使用之自動轉向系統。 研究方法包含了以二維單軌模型建立車輛自動化轉向模型及路徑追蹤模型狀態方程式以了解系統之特性；實際硬體以Ez-Guide 250系統搭配Bulkhead天線做為GPS接收端進行路徑規畫及導航，搭載Ez-Steer 500系統控制器及Ez-Steer Motor致動馬達做為自動轉向系統之控制單元及致動器操作農業機械方向盤，以模擬人為操作方向盤動作進行自動轉向功能。將此系統裝配於台灣特有之一型農業機械─桃改型桿式噴藥機SYL-800進行自動轉向運作測試，並以Matlab程式撰寫一資料後處理程式。 實驗以APP程式GPS test先行確認實驗場地之GPS訊號良好，之後於實驗場地進行七輪實驗，分別測試噴藥機以定速2.3 km/h行走時追蹤直線、直角轉彎及連續轉彎；接著是測試噴藥機定速3.6 km/h行走時追蹤兩組直線、直角轉彎及迴轉，其中兩組直線第一組設定參數與定速2.3 km/h時設定相同，而第二組則為符合定速3.6 km/h之優化參數，以此方式驗證模型表現之特性。 實驗結果顯示在參數條件設定得當的情況下，系統可以進行良好的自動轉向功能進行路徑追蹤。以定速2.3 km/h或是3.6 km/h進行直線路徑追蹤都可以達到行走誤差極小可忽略的良好路徑追蹤，而直角轉彎則可以在誤差小於0.5 -1.6 m以內的情況下完成。然而在進行連續轉彎或是迴轉時，系統可能會因為控制指令變化過於不連續導致系統跟不上而中止運作，需於系統應用時注意此局限。

In the age of automated industry, agricultural machines are also becoming more automated. This paper implements a GPS-guidance system that allows Taiwanese agricultural machinery to be steered automatically. The study is in two parts. Firstly, a single-track vehicle model is used to understand the characteristics of auto-steering and path-tracking. A GPS guidance system is then implemented in a field test. An Ez-Guide 250 monitor with a GPS receiver and a bulkhead antenna are used as the guidance system. The Ez-Steer 500 system includes a control unit and a driving motor that imitates a human’s actions in turning a steering wheel to allow auto-steering. The system is then installed on a self-propelled field boom sprayer, named SYL-800, which is manufactured in Taiwan, to execute the field tests. A computer program is then written using Matlab, to analyze the data recorded from this system. The experiments initially test the GPS signal using an APP program named GPS test, to ensure that the signal is sufficiently strong for the system during the field test. Experiments are performed in seven rounds, as follows. During the first three rounds, the speed of the sprayer is fixed at 2.3 km/h and the system tracks a straight line, a cornering line and a continuous turning path, respectively. During the next four rounds, the system tracks straight line, cornering line and U-turn line trajectories and the speed of the sprayer is fixed at 3.6 km/h. During these four experiments, the straight-line path is run at two different factor settings: the sprayer’s speed is set to 2.3 km/h and 3.6 km/h. This method is used to prove the characteristics of mathematical model that is established. The experiment results show that the system works well as an auto-steering system for path tracking when there are proper factor settings. The system tracks a straight line with negligible deviation regardless of whether the speed of sprayer is 2.3 km/h or 3.6 km/h. The deviation in the cornering line is within 0.5 -1.6 m. However, the system shuts down when the path is a continuous turning line or a U-turn line, because the control signals are not continuous. Care should be exercised when running the system in this mode.