全國中小學科展

2021年

Biodegradation of Post-Cured Photopolymeric Resin of Stereolithography 3D Printers Using Galleria mellonella Larva.

The present research has as main objective to degrade the post-cured photopolymer of the stereolithography 3D printer resin using Galleria mellonella larvae. It is necessary to consider that the use of materials from 3D printers tends to increase considerably and in approximately seven years about 10% of everything that will be produced in the world will come from this type of printing. Considering also that the increase in population growth and technological development are directly linked to the increase of solid waste on the planet, in particular to polymeric materials, there is a need to degrade and give an adequate end to waste, avoiding a notorious accumulation along the time. For this purpose, Galleria mellonella larvae will be used because of it's comprovated capacity to degrade polyethylene, to find out if it is capable of biodegrading the post-cured resin of the printer. To carry out the research, compositional tests were done in partnership with the SENAI Institute for Innovation in Polymer Engineering, located in São Leopoldo, Rio Grande do Sul, and the creation of the larvae and degradation of the photopolymer will be carried out in partnership with the University Federal University of Health Sciences of Porto Alegre (UFCSPA). The data analysis will be based on the crystallinity determination tests by differential scanning calorimetry (DSC), thermogravimetric analysis (TGA) and attenuated total reflectance spectroscopy (ATR) that will also be applied in the larvae feces after contact with the polymer to assess for degradation. As a result of the compositional tests, the ATR showed predominantly characteristic absorptions of acrylic resin; in the TGA test, the loss of mass described in the test is related to the loss of mass of organic material, mainly polymer. Finally, in the DSC test a thermal event was observed in the heating of the sample, with peaks at 125 ° C (Tpm), characteristic of fusion, and a thermal event in the cooling of the sample, in 112 ° C (Tpc), characteristic of crystallization. Based on the analysis of the results obtained, it is possible to infer that most of the composition of the photopolymer is acrylic resin, widely used in stereolithography 3D printers. The research has the future objective of isolating the substance into the larvae responsible for degradation so that it can be degraded on industrial scales. The research started in March 2020 and is still under development due to the COVID-19 pandemic, which compromised the planned tests.

中文重點文句摘取

在資訊爆炸的時代,效率閱讀、整理資料的能力越趨重要。身為高中生,學習時的閱讀量龐大,還須另外自己挑選重點句,重新整理筆記。因此我想如果可以讓電腦自動摘取文章的重點,就能幫助學生效率學習。 大多數現存的自動摘要研究適用於英文文本,本研究利用演算法抓取中文文章的摘要,使學生可以真正實用該演算法於日常學習當中。除此之外,此研究比較了不同方法摘要的準確率以及優缺點。

探討紅綠色盲族群對不同顏色赫曼方格的視錯覺現象

延續2020年視錯覺研究,我們想要探討視覺神經細胞接受不同亮度對立顏色刺激後是否會有不同活化程度,影響視錯覺?另外也想探討不同對立配色的赫曼方格中線條寬度的閾值是否有差異?我們設計網頁版程式,改變赫曼方格線條顏色的亮度和寬度,以正常受試者為對照組,色盲受試者為實驗組,探討視錐細胞的變異是否影響視錯覺的產生。我們以顏色拮抗理論定義對立配色的赫曼方格,以標準化情境進行實驗。在線條亮度的結果中,黑白配色的赫曼方格圖案中線條與方格對比度越大,視錯覺產生的效果越明顯。兩群受試者在綠紅、紅綠、藍黃這三組配色,線條亮度範圍呈現兩極化的趨勢,可能原因為色盲因錐狀細胞變異,需要較高亮度的對比顏色刺激才容易讓神經細胞有活化程度上的差異,造成視錯覺出現。在改變線條寬度的結果中,發現藍黃的赫曼方格閾值與明顯值最大,黑白次之,綠紅最小,可能與細胞感受野大小有關,且線條寬度超過感受野的閾值,視錯覺就無法產生。兩群受試者的線條寬度閾值差異,可能與色錐細胞變異有關。綜合以上結果,顯示視網膜色錐細胞與神經節細胞的作用,對視錯覺形成扮演一定角色。在消費行為問卷結果中,我們發現格子衣服若是顏色對比度大、線條較細的圖案設計,消費者的滿意程度較低,這原因可能與視錯覺產生影響衣服視覺效果。

懸浮微粒三維偵測與預報系統

近年空氣品質已是居住環境與健康的指標,「細懸浮微粒」充斥在空氣中,造成過敏,增加肺癌的危險。本研究探討懸浮微粒在受到重力、空氣阻力與空氣浮力影響後,形成分層。並利用VPython軟體模擬不同大小的懸浮微粒(pm10、pm2.5與pm1.0)於空間中碰撞及受到空氣阻力產生能量衰減,藉此了解不同微粒之分層現象。再實作以居住樓層不同的垂直高度,設計組裝架設「懸浮微粒三維偵測器」及物聯網。以台灣中部地區,日益增加的空汙狀況下,模擬以台中火力發電廠為例,探討其風向、地理環境、以擴散模式理論模擬後,選定數棟建築物,監測每棟建築物地面上不同高度的空氣品質數值。最後監測數值自動上傳至物聯網雲端資料庫 ThingSpeak,並可於使用者端監測及取得測量數值;期許再利用機器學習及歷史累積的三維空氣品質資料,將來更優化預測空氣品質數值之成效。

Detect the Defect

"When the Well is Dry, we will know the Worth of Water." Most of 埃及 and the world suffers from water and petrol shortage. With the current consumption rate, two-thirds of the world's population may face water shortages by 2025. These are water pollution, overpopulation, and agriculture, leading to wastewater from landfills and pipes that seep into the ground and may pollute the water, making it unfit for human consumption and waste more water. Besides, some accidents happen to water distribution and irrigation systems, causing a significant loss in water. According to the ministry of water resources, in 2016, the need for freshwater is 67 billion cubic meters. On the other hand, 埃及 receives only 55 billion cubic meters (2.6 billion cubic meters of them evaporate during runoff). Also, one of the wasting water methods in modern irrigation systems is water leakage from pipes as the water transmission and distribution lose about 31% of the produced water due to pipe leakage. Besides, every day more than 3.3 billion liters of treated water – 20 percent of the nation's supply and 234 million liters a day more than a decade ago – are lost through leaking pipes in England and Wales. Many reasons lead to leakage in pipes like water pressure, clogs, and corrosion. The leakage in pipes does not exist in the lines of water only. Also, the pipes in a petrol can cause dangerous accidents like the accident in the Bahira government that led to the death of 6 people and made 19 in a dangerous state. Our project designed a system that can detect fluid leakage and deal with it fast to prevent the wasting of fluid by using sensors and electronic circuits. Our system provides us with information about the fluid (like the amount of the flowing fluid and its speed). Therefore, if there is a difference in the reads, we understand that there is a leakage in this region, and the system will automatically stop the fluid flowing through the pipes. the system will locate all the leakage sites and send them to the mobile app with the amount wasted and the actions taken.

探討粒線體對果蠅卵巢生殖幹細胞維持的影響

幹細胞會進行不對稱分裂以維持組織恆定,一個子細胞分化為有特定功能的細胞,而另一個則維持其細胞潛能。粒線體能進行分裂和融合並維持動平衡。現今的研究已經了解,誘導型多能幹細胞(iPSC)的粒線體動平衡變化會促使其直接分化 (Seo BJ, et al., 2018)。然而,果蠅卵巢生殖幹細胞(GSC)內粒線體動平衡與幹細胞分化的關聯仍不清楚。本研究利用容易辨認的果蠅卵巢生殖幹細胞來探討這個問題,並使用Gal4-UAS系統與RNAi操縱粒線體動平衡,並發現粒線體融合蛋白被抑制會導致油滴堆積及幹細胞損失。另外,使用左旋肉鹼(L-carnatine)增進脂肪酸代謝,發現增進代謝會導致油滴減少及幹細胞回復。本研究著重於探討粒線體動平衡對GSC維持與脂肪酸代謝的影響,並對粒線體脂肪酸代謝與幹細胞分化的潛在關係提出觀點與佐證,盼未來能有更進一步的研究與醫療方面的應用。

基於深度學習之服裝試衣系統

本研究以AI虛擬試衣系統(Virtual Try-on)為主題,透過深度學習技術,並結合幾何匹配模型,開發出試衣系統,可將使用者上傳的照片,模擬成穿著新衣的模樣。 首先,以深度學習模型將人物原始圖片取出骨架節點,並生成人體遮罩以及保留人物頭部,再結合以上三種資訊合成為高維特徵圖。接著將目標替換衣物生成出依照人體姿態扭曲後的衣物圖片。最後於Virtual Try-on模型中將人體高維特徵圖與扭曲衣物作為輸入,並經過深度學習網路合成出穿著目標衣物之人體圖像。本研究結果發現,人物站姿單純,且雙手緊貼身側,以及拍攝角度為正面、衣服款式為短袖、背景色彩對比度較高與衣服圖案單純的原始圖片,可得到較好的合成結果。

Ecological inks for markers

Markers have become essential in school and work life due to their great usefulness for teaching and homework. Despite the benefits they have brought, markers are the cause of great contamination from the ink manufacturing process to the excessive production of plastic. Ecological inks in markers and the innovative design of a refillable marker, allow to generate less pollution without having to stop using this product. From dyes created with coffee, fruits and vegetables that pass through different processes, natural inks arise that replace the use of polluting dyes. Likewise, implementing recycling plans in different institutions, markers that were no longer used were collected to be filled with ecological inks and used again. In addition, the excessive production of plastic is reduced by selling and refilling markers and ink kits.

以結膜影像判斷貧血之研究

全球貧血人口普遍,然許多人並不了解自身是否罹患貧血;長期患有貧血的病人,亦需定期抽血檢驗追蹤是否有貧血惡化達到需接受輸血的程度。研究顯示,結膜之顏色與貧血有絕對關係,結膜越白則貧血越嚴重,醫師也常使用結膜顏色推測是否有貧血情形。若能設計手機軟體自動分割結膜影像並分析其顏色,將有機會推測受試者是否罹患貧血。本研究收集22位無貧血者及8位貧血病人,並獲得其近期血紅素數值。以手機取得受試者之眼睛影像後,成功設計程式以深度學習完成結膜自動影像分割,對於分割影像以面積大小進行後期處理後,依其取得下眼瞼結膜之三原色平均,再利用kNN與SVM演算法判斷預測出該受試者是否具有貧血之症狀。本研究主要分為兩階段,其一為進行下眼瞼結膜分割模型訓練;其二為製作有無貧血之判斷模型。整合上述眼瞼分割模型(IoU=89.8%±0.02%)與貧血判斷模型(SVM以polynomial核函數測出 準確值93.3%±24.3%)後,可得貧血診斷準確率為80%。此結果代表AI技術有機會透過結膜影像,判斷被拍攝者是否有貧血情形,未來若能增加研究人數,將可設計網頁版或手機APP加以推測血紅素值,供大眾居家篩檢。

THE DESIGN OF MICROFLUIDIC PUMP (MFP) FOR MEDICAL FIELD

The ability of microfluidic (MF) device technologies to provide a lot of information with a small amount of sample, the opportunities it offers increases their use in the medical field in the bedside monitoring in drug delivery systems. Three-dimensional (3D) printer technologies provide advantages such as cost-effectiveness in the production of MF devices and quick and easy production in intricate designs. In our project, it is aimed to design microfluidic pumps (MFP) to be used in the medical field and conduct its production with 3D printer technologies. The developed MFP is intended to be at low cost, bio-compatible, adaptable, and portable to the drug, suitable flow properties as a pharmaceutical pump. First of all, MFP air channel, flow channel, etc. parts were designed and printed with the help of a 3D printer and on AutoCAD, one of the professional drawing programs. The poly(dimethylsiloxane) (PDMS) membrane that will enable MFP activation is produced in different thicknesses and glued to the air channel of MFP. The resistance to the applied pressure is observed, and the appropriate membrane thickness is determined as ~ 235µm. Liquid PDMS was applied to the inner surfaces of MFP's air and flow channel, PDMS membrane was placed between them, and the parts were assembled in the oven at 60ºC. MFP has been connected to the pneumatic valve system, where operation codes have been prepared with Arduino Uno, and flow properties have been examined. The flow rate of MFP is ~ 50 µL/min at a maximum of 15 Hz, and the backpressure is ~ 0.085 Pa under a maximum pressure of 3 bar. Also, values such as size, membrane thickness, and applied pressure for the possible models of MFP were supported by theoretical calculations. As a result, MFP, which is biocompatible, drug adaptable, portable, wearable technology application potential, and has suitable flow characteristics as a pharmaceutical pump, has been developed. MFP introduced a microfluidic pump system that can make life easier for the patient and contribute to the national economy through domestic production and can be used as a drug pump in the treatment of diseases such as diabetes and cancer.