本研究為車管立把之FPGA即時影像自動化取放系統,即利用FPGA影像系統進行物件角度辨識,並配合機電整合系統以完成自動化取像。研究內容主要有機器手臂結構應力應變分析、FPGA影像辨識、伺服馬達最佳化參數設計及PLC整合控制實驗。實驗結果顯示機器手臂支架承受本身重量負載時,其最大應力、應變與位移的值分別為0.853 N/mm2、8.597×10-6、3.511×10-2 mm;當機器手臂支架承受至98.1N的力量時,其承受的最大應力只有8.26 N/mm2且最大變形量為0.3165mm相對於機器手臂支架本體來說影響並不大;機器手臂支架的自然共振模態頻率分析中未受限制機器手臂支架的第一個自然頻率為32.60 Hz,其震動方向為繞著支架下方固定端進行前後方向轉動。FPGA不同角度檢測誤差實驗中,量取的角度與實際偏移角度誤差最大為1.1度,其餘的角度量取誤差皆在0.6度以內,且平均誤差值為0.3523度。線性滑軌速度過快時容易造成線性滑軌衝出感測器的感測範圍,而當速度到達11000 mm/min時,線性滑軌會因無法承受高速而產生暴衝及線性滑軌不穩定等現象,故建議最佳移動速度為7000到9000 mm/min之間。手臂夾爪各旋轉角度重複性中,角度的實際量取誤差值與理論換算值間的誤差大概都在1~3度左右,而誤差量取的最大及最小值在角度越大時,會有較小的誤差趨勢。
This study designs the automatic pick and place system of stem by real-time image base on FPGA. The main purpose of this automatic manufacture is using FPGA image processing to recognize the angle of a bicycle stem. The mechatronics system consists of three servo motors and motor driver, a sensor, a field-programmable gate array (FPGA), a stepper motor, a conveyor belt, a linear slider, four pneumatic cylinder and a robot arm mechanism. Experimental results showed that the values of the architecture stress, strain and deformation are 0.853 N/mm2, 8.597×10-6 and 3.511×10-2 mm, respectively. When a load force of 98.1N is applied on the bracket, the maximum stress and deformation are 8.26 N/mm2 and 0.3165mm, respectively. The stress and deformation of the bracket are not affection. The first natural frequency analysis and vibration directions of the bracket are 32.60 Hz and turn around the center of fixed end of the bottom before and after directions. In the measurement of different angle detections, the maximum angle error is 1.1 degrees and the rest of the angle errors are within 0.6 degrees. In the speed setting of a linear slider , the sensor can’t sense the original reset signal when the linear slider moves too speedy. The input speed of linear slider cannot be set bigger than 11000 mm/min, because the linear slider will easily produce shock. The error difference between theoretical and practical of rotation repeatability is 1 to 3 degrees.