| Summary: | 碩士 === 國立中興大學 === 農業機械工程學系 === 89 === The most eel cultivator in Taiwan uses the traditional bar-tray sorter to grade the eels at present. The machine is very accurate, but it is time consuming and labor costing. The purpose of this research is to develop a new eel sorting machine which is expected to achieve accurate grading for the counting of the elvers and grading of eels. It is also looked forward to save the labor power and accelerate the grading speed so as to promote the development of the eel industry.
The grading mechanism of the new eel sorting machine applies two bar-trays, one in above and another in below. The former is fixed, and the later is adjustable to change the grading spacing by moving up and down. After finishing the grading of the smaller size eels, the next grade can be obtained by increasing the grading spacing. Therefore, the usually occurred blocking problem of the traditional bar-tray sorter is solved, and the performance of grading several grades using the same sorting machine is also achieved. The cylindric bar is selected in this mechanism, so the realistic grading spacing is only the point of tangency between two bars. However, eels need to go through longer distance to pass through the grading spacing in the traditional bar-tray sorter with the use of rectangular bars. As compared to the traditional sorter, the new developed mechanism also causes lighter damaging to eels during the grading process.
After the test of the grading precision, the mean grading precision of the prototype machine is 55.13 %. It shows that after grading, there are 55.13 % eels in the acceptable weight range. According to this result, a deviation in the spacing settings of the prototype machine is inferred to happen. This causes error of the grading work resulting from the large grade eels falling down the small grade. After measuring the setting spacing of the prototype machine, there is 53.545 % of the setting spacing in the ±5 % range of the standard value. It represents the spacing accuracy of the prototype machine is 53.545 %, influencing the grading precision.
The test of the grading precision of the single grading path to further confirm the precision of this grading mechanism, the mean grading precision of 77.52 % is obtained, better than the prototype machine. Using the statistics regression method to analyze the data obtained during the test, an exponential equation is constructed to express the relation between the eel size and the grading spacing. The coefficient of determination R2 of the equation is 0.9927, better than the R2 0.9797 of the traditional sorter.
Furthermore, the test of conversion from the gross weight into the fish number, the accuracy of the prototype machine is 93.54 %. There are 94.44 and 92.99 % for the twice tests of the single grading path, respectively closing to the accuracy of the prototype one. It proves that the accuracy of this grading mechanism on the calculation of fish number using the conversion from the gross weight is higher than 90%.
|
|---|
