自主式養蝦投餌機器人之研發

Development of an Autonomous Shrimp Feeding Robot

蘇奕綸(Yi-Lun Su)

指導教授 ： 朱元南

國立臺灣大學/生物資源暨農學院/生物產業機電工程學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.00616

摘要

本研究之目的是研發一個用於蝦類養殖的自主式投餌機器人。養殖時每日的投餌工作是一項基本卻重要的程序，一般投餌時主要之注意事項為控制投餌時間、次數與投餌量等。而為了配合蝦子的生態習性，於夜間進行少量多餐式的投餌是較佳的投餌方式；在投餌量控制方面，投餌時須配合蝦子的不同成長期以及氣候與水質狀況等因素，來判斷投餌量，若投餌過量或過少都會影響養殖成效。但現實中，目前傳統的人力投餌方式難以達成以上幾點條件，因此本研究設計一個投餌機器人，讓此機器人可於夜間自主式的進行少量多餐的投餌，以改善投餌方式，增進養殖成效。
為使機器人能在養殖池上行動投餌，故此機器人決定採用「機器船」的形式，並使用船外機作為機器人前進與轉向之推進系統。行動模式方面，因蝦子之習性是遍佈在養殖池的底邊四周覓食，所以設計使機器人以沿著養殖池邊約1.5公尺處，繞行養殖池數圈的方式進行投餌。利用超音波感測器偵測機器人與岸邊之距離與角度，再利用模糊控制法則依此資訊計算機器人所需的轉向控制，以達成機器人的沿牆行走(Wall-following)控制；實驗結果顯示，機器人可與沿牆目標保持在1.5公尺±20%的距離之間前進，足以達成養殖投餌應用之需求。接著在停靠點設置紅外線發射器，使機器人投餌結束後，能利用自身之紅外線感測器，感測停靠點位置後返回停靠。機器人所使用之轉盤式投餌機構，投餌範圍包含半徑1.5公尺之範圍，並可依照投餌需求的不同，利用改變船速配合調整投餌頻率的方式，使機器人完成繞行養殖池四周的投餌量控制。

關鍵字

水產養殖 ； 蝦 ； 投餌 ； 機器人 ； 模糊控制

並列摘要

This research focused on the developing of an autonomous shrimp feeding robot for improving the feeding method and cultivation results. Feeding shrimps day by day is a basic but important work in cultivation period. The feeding time, amount and frequency are the major factors that affect the results most. It is a better way to feed less more often at night for conforming it to the behavior of shrimps. For the amount of feeding, it depends on the weather, season, water quality, and the phase of life cycle in shrimps etc. No matter we overfeed or underfeed, it might reduce the cultivation results. In tradition, it is difficult to achieve the best feeding way using manpower. Instead of that this research developed a feeding robot and applied it to feed less more often at night.
Because the robot feeds and acts on the surface of aqua water, it was designed as the type of a boat. For the function of moving, it was made of a transom-mount trolling motor. Since shrimps usually forage along the edge of the bottom at a breeding pond, the robot was designed for moving along the banks several times in order to feed shrimps. And it kept a 1.5 meter distance away from the bank.
Ultrasonic distance sensors were used on the robot to determine the distance and angle between banks and itself. According to those data, fuzzy logic control made the robot accomplish wall-following action. The results indicated that the robot could move along the bank and keep a 1.5±20% meter distance away from the bank. An IR transmitter was built up in the berth. The robot could rely on the IR receiver mounted on it to determine where to berth itself after finishing the job. The feeding mechanism was swing the bait out with a turntable, and it could feed within a radius of 1.5 meter. The mechanism could adjust the speed of the robot or the frequency of swing the bait to control different kinds of feeding amount.

並列關鍵字

Aquaculture ； shrimp ； feeding ； robot ； fuzzy logic control

參考文獻

2.戴天適。2004。使用傾斜儀及加速計之指向裝置的定向控制。碩士論文。台北：國立台灣大學電機工程系。

3.Bank, D., and T. Kampke. 2007. High-resolution ultrasonic environment imaging. IEEE Transactions On Robotics, 23:370-381.

4.Barshan, B. 2001. Ultrasonic surface profile determination by spatial voting. In “Proc. 18th IEEE Instrumentation and Measurement Technology Conference, 2001”, 1:583-588, Budapest, Hungary.

6.Chung, W., C.B. Moon, K.R. Kim, and J.B. Song. 2006. Design of a sensor model and semi-global localization of a mobile service robot. In “Proc. International Joint Conference on SICE-ICASE”, 4260-4265, Busan.

7.D’Andrea, B., and K.A. Abderrahim. 1995. Control of nonholonomic wheeled mobile robots by state feedback linearization. International Journal of Robotics Research, 14:543-559.