Chung Cheng University Institutional Repository
Not a member yet
889 research outputs found
Sort by
中醫方劑配伍系統;A Formula Combination System for Traditional Chinese Medicine
No full text[[abstract]]本研究團隊已研製一個中醫辨證系統及一個中醫藥材配伍系統。本論文在前兩個系統的基礎上，研製一個中醫方劑配伍系統。這三個系統實踐中醫「辨證論治」的資訊化。中醫辨證系統實踐中醫「辨證」的資訊化，目前支援67個證候。中醫方劑配伍系統及中醫藥材配伍系統協力實踐中醫「論治」的兩階段資訊化。第一個階段，中醫方劑配伍系統根據辨證的證候，推薦治療效力最強的方劑，目前支援186帖方劑。第二個階段，中醫藥材配伍系統根據辨證的證候及推薦的方劑，進行加減藥材，推薦治療效力最強的藥材，目前支援334味藥材。本論文也進行新增藥材及方劑的標準化，並依標準化的藥材及方劑更新中醫藥材知識本體及中醫藥材辭庫。本論文也訂定各方劑治療各證候的效力值，運用多目標最佳化技術，推薦治療多證候效力最強的方劑。
Our research team has developed a syndrome differentiation system and an medicinal combination system for traditional Chinese medicine (TCM). This thesis developes a formula combination system for TCM based on those two systems. These three systems apply the information technology to implement the framework of “syndrome differentiation and treatment” in TCM. The syndrome differentiation system implements the portion of “syndrome differentiation” and supports 67 syndromes currently. The formula combination system and the medicinal combination system collectively implement the portion of “syndrome treatment” using a two-stages approach. At the first stage, the formula combination system recommends the most effective formulas according to the syndromes identified by the syndrome differentiation system. It supports 186 formulas currently. At the second stage, the medicinal combination system allows practitioners to further add most effective medicine to or delete medicine from the selected formula. It supports 334 medicines currently. This thesis also standardizes new medicines and formulas, and updates the medicinal ontology and medicinal thesaurus accordingly. This thesis also sets the effectiveness of formulas for treating associated syndromes, and applies the multi-objective optimization technique to recommend the most effective formulas for multiple syndromes
基於深度強化學習之自動駕駛車輛路口控制系統;Intersection Crossing for Autonomous Vehicles based on Reinforcement Learning
No full text[[abstract]]在路口發生的交通事故將會嚴重影響交通流量，且該類型的交通事故佔有很大的發生比例。在人口與車輛數量均增加的趨勢下，發展更安全而且有效率的路口控制系統有其必要性。而且隨著車載通訊與自動駕駛技術的進步之下，我們可以發展更聰明的控制系統，可以減少人為因素所造成的種種交通意外。 因此，我們提出了基於深度強化學習之自動駕駛車輛路口控制系統(Deep Reinforcement Learning-based Autonomous Intersection Management，簡稱DRLAIM)。在這個系統中包含三個主要模型，分別為安全煞車控制模型、基於強化學習之路口控制模型及優先權分配模型。我們希望透過強化學習去訓練我們的系統，讓其在與交通環境的互動中學習到一個好的交通控制策略，並且使用安全煞車控制模型去確保每一輛自動駕駛車輛在行進中的安全。 實驗結果顯示，我們的路口控制模型在使用強化學習訓練之後，道路的吞吐量增加了83%。與快速優先服務策略(Fast First Service，簡稱FFS)相比，DRLAIM的車輛平均等待時間較少。減少的幅度分別為1.2%到11.4%。
Road accidents at intersection significantly impact traffic flow and account for a significant proportion of all accidents. With the trend of increasing population and vehicle numbers, it is necessary to develop a safer and more efficient intersection control system. Furthermore, with the advancement of vehicular communication and self-driving technology, we can develop a smarter control system that can reduce the traffic accidents caused by human factors. For increasing safety and efficiency of traffic environment, we propose a Deep Reinforcement Learning-based Autonomous Intersection Management (DRLAIM) system. This method contains three main models, including brake-safe control model, intersection control model based on reinforcement learning and priority assignment model. We train the system to learn a good intersection control policy by interacting with traffic environment through reinforcement learning and the brake-safe control model is used to ensure the safety of each autonomous vehicle during travel. Experiment results show that after training using reinforcement learning, the throughput of intersection control model increased by 83%. In comparison with the Fast First Service (FFS) policy, the average waiting time of DRLAIM reduced by about 1.2% to 11.4%
石墨氈電極壓縮率對全釩氧化還原液流電池性能影響之研究;Study on the Effect of Compression Ratio of Graphite felts on the Performance of All-Vanadium Redox Flow Batteries
No full text[[abstract]]能量儲存系統（ESS）在綠色能源技術的發展中發揮關鍵作用，因為再生電源（例如太陽能和風力）的固有間歇性和不穩定的功率輸出。在大型能量儲存系統中，全釩氧化還原液流電池（VRFB）被認為是有前景的ESS，由於其擁有靈活的功率規模設計、高周期壽命、和高安全性。 影響VRFB性能的因素包括質子交換膜，多孔電極，電池設計和操作條件。由於電化學反應發生在質子交換膜和多孔電極中，因此VRFB的性能深受這兩種材料影響。在這項研究中，比較和討論質子交換膜和石墨氈的酸處理條件對效率的影響，然後再研究石墨氈的壓縮率對VRFB性能的影響。我們測量和比較各石墨氈壓縮率的VRFB單電池阻抗，在各種操作條件下。最後，設計一個特別夾具用於測量VRFB的局部電流密度（LCD）分佈的變化。充放電期間，我們測量和比較在各石墨氈壓縮率和操作條件下的LCD流動方向。 結果表明，24 h和12 h的質子交換膜和石墨氈的酸處理表現出最佳能效86.42 %。此外，高頻電阻隨著石墨氈的壓縮率的增加而降低。LCD在充電期間沒有顯著變化。對於較高壓縮率的石墨氈，LCD的充電狀態不均勻性明顯較低。
Energy storage systems (ESSs) play a key role in the development of green energy technologies because of the inherently intermittent and unstable power output of the renewable power sources, such as solar and wind power. Among large-scale energy storage systems, all-vanadium redox flow batteries (VRFBs) are considered as promising ESSs because of their flexible power scale design, high safety, high efficiency, deep charge, and long cycle life. There are many factors affecting the performance of a VRFB, including materials of the separator, porous electrodes, bipolar plates, battery design, and operating conditions. The electrochemical reaction occurs in the separator and porous electrodes, therefore, the performance of the VRFB is affected by the properties of these two components. In this study, a proton exchange membrane and graphite felts are treated by sulfuric acid and the effect of treatment conditions on the efficiencies are compared and discussed. Then, the effect of compression ratio of the graphite felts on the performance of the VRFB are investigated. The impedance of the single VRFB cell with various compression ratios of graphite felts at various operating conditions are also measured and compared. Finally, a specially designed fixture is employed to measure the variations of local current density (LCD) distribution of a VRFB. The LCD distributions along the flow direction at various compression ratio of graphite felts and operating conditions during charge and discharge are measured and compared. The results suggest that the VRFB with a separator and graphite felts treated by sulfuric acid for 24 and 12 h, respectively, demonstrates an optimal energy efficiency of 86.42%. In addition, the high frequency resistance decreases with increasing compression ratio of graphite felts. The LCD shows no significant variation during charge. The state of charge where the LCD becomes dramatically non-uniform is lower for a higher compression ratio of graphite felts
自發性熱聲引擎之聲轉熱動態分析;Dynamic Thermal Analysis of Self-excited Thermoacoustic Engine
No full text[[abstract]]本論文主要是利用數值模擬，引入一個迴授系統的觀念來設計自發性的熱聲引擎。熱轉聲跟載具的頻率，可以藉由管長與載具的性格來推導出來；但多孔性材料的頻率響應卻無法得知，因此利用數值模擬軟體，來研究多孔性材料聲轉熱的頻率響應，得知頻率響應後畫出系統的波德圖，就能知道熱聲引擎的最佳工作頻率來進行設計產生不穩定，也能用來判斷所設計的熱聲引擎是否有效率。
In this thesis, we use numerical simulation and introduce a feedback system to design a self-excited thermoacoustic engine. The frequency of thermal transfer and carrier can be deduced from the character of tube length and carrier. However, the frequency response of porous material can not be known.Therefore, the numerical simulation software is used to study the frequency of acoustic heat transfer Responding to the knowledge of the Bode plot of the system after the frequency response shows that the best working frequency of the thermoacoustic engine is known to make the design unstable, it can also be used to judge whether the designed thermoacoustic engine is efficient or not
二氧化碳在碳酸化過程中最佳化機制研究;Study on Optimization of Carbon Dioxide Mechanism in Carbonated Process
No full text[[abstract]]本研究探討的是二氧化碳在常溫溶液中碳酸化過程機制的研究，一般市售氣水製程採用的碳酸混合機為先冷凍降溫水溶液再灌入二氧化碳氣體進行混合，此研究以常溫水溶液為主，灌入二氧化碳氣體讓其進行混合，並記錄其灌壓壓力、混合中壓力、氣體溫度、混合完時溶解度、裝瓶後溶解度，探討其關係，有學校實驗與廠商試驗機測試。學校實驗，由於經費與場地大小的限制關係，採用高壓罐灌壓實驗，並記錄相關數據。 配合廠商的混合機試驗機測試，參考學校實驗結果設置參數，進行試驗並記錄其測試結果，以在常溫溶液下進行碳酸化並達成達成產品目標
微型癌細胞阻抗量測模組之研發;Development on the Miniaturized Impedance Measurement Module for Cancerous cells
No full text[[abstract]]分析細胞數量與細胞型態提供了我們在早期判斷癌症的重要依據，並即時地監控療程進度。利用電性檢測的原理在近幾年成為非常熱門的技術。在文獻中透過阻抗量測來檢測癌細種類的技術已經備受關注。本文提出了一種利用介電泳阻抗量測（DEPIM）方法研究人肺細胞株的微流體晶片，圓形與直線形狀的指叉式微電極的微電極圖案是由微製程技術製作而成。分別將不同的細胞數樣本注射在次毫米體積的微腔室中，其中細胞不僅能透過介電泳增強其集中在微電極表面上的現象，更可從捕獲在微電極上的細胞測量出電阻抗特性，並更進一步地發現在每種細胞的導納值變化與細胞數之間存在著線性關係。藉由這種特性可以量測出人類癌細胞(A549)與正常細胞(MRC-5)之差異，進而達到癌細胞檢測之目的。
Cancer has become a worldwide disease, and increases rapidly in recent years. A precise treatment depends on the accuracy of the diagnostic technique, in which includes analysis of cell number and cellular morphology. This study presented the development on the microfluidic device utilizing the dielectrophoretic impedance measurement (DEPIM) method which explores the human lung cell lines. The microchip consists of interdigitated microelectrodes with the circle -on-line shape and a micro chamber in which both DEP concentrating and impedance sensing for cells inside. This micro device has been fabricated by standard microfabrication processes. A highly sensitivity impedance measurement circuit board was built for the chip. It provides the sinusoidal signals with the frequencies in a wide range, so that the experiments can be performed in a reliable excitation. The results show a linear relationship between the admittance variation and the cell number in the different cell lines. The significant difference in the slope of these characteristic lines were adopted to distinguish two different lung cells, including A549- human lung carcinoma cells and MRC-5- human normal lung epithelial cells. The device is a simple-to-operate with high sensitivity, rapid response, inexpensive and portable application for cancerous cells
由EVA原料預估發泡材料在不同壓縮速率下材料性質評估;Estimation of Material Properties of Foamed Materials From EVA Raw Materials at Different Compression Velocities
No full text[[abstract]]發泡材料普遍地應用在現今人類生活中各個領域包含防護器具、建築工程、運輸包裝等，其功能主要為吸震與緩衝，並用以保護物體結構或人體免於受損傷害，而本研究則探討用於運動鞋中底的材料乙烯/醋酸乙烯酯共聚物(Ethylene Vinyl Acetate，簡稱EVA)製程，發泡材料由高分子原材料、發泡劑等多項配方混合後，經由射出成型機射出發泡成型。而在過去的製程中並沒有一套有效率的方法藉由材料的配方判斷發泡材料的材料機械性質，因此本研究建立一套預估公式，藉由發泡材料的配方並配合實驗的製程參數預估發泡材料的材料機械性質。本研究分為實驗與理論公式兩大部分，實驗的部分內容包括射出成型發泡實驗以及靜、動態壓縮實驗，從實驗結果得到製程參數對於公式內係數的對照表。另一方面理論公式是整合Mills和Gibson兩位學者的假設模型，讓發泡預估公式可以完整描述發泡材料的材料機械性質，由此提供將來在工程上的應用。
Nowadays, foam materials are widely used in various fields of mankind life like vehicle engineering, construction engineering, packaging materials and so on. The main functions of this type of material are shock absorbing and impact cushioning, so it could be used to protect the structure or human body from injury. This study focuses on the material which shoe midsoles are usually made of called Ethylene Vinyl Acetate. In the past, there was no efficient way to determine the material mechanical properties of foamed materials by the formulation of materials. Therefore, this study established a set of predictive formula, by the raw materials and the experimental process parameters. This study consists of two major parts. The first part is experiment including injection molding and quasi-static and dynamic compression test. From the experimental results, a control table of experimental process parameters versus formula coefficients is obtained. On the other hand, the theoretical formula is to integrate the hypothetical models of Mills and Gibson scholars, so that the foam predictive formula can completely describe the foam material mechanical properties, thus providing future applications in engineering. The results from the study could be used in the engineering application
應用內嵌感測器於工具機滑塊智能化之研發;Sensor-embedded Linear Ball Bearing for Linear Guide Way Straightness and Parallelism Monitoring
No full text[[abstract]]本研究主要發展智能化滑塊，將感測器內嵌於線性滑軌之循環式滾珠滑塊中，此滾珠滑塊中安裝兩種感測器，第一個為提供定位訊息的光學位置傳感器，第二個為觸覺式秤重傳感器，稱為薄膜式壓力傳感器(Force Sensitive Resistor, FSR)，也稱為力敏電阻，此感測器量測來自導軌的力量，利用這兩種感測器檢測滾珠滑塊在線軌上的位置以及線軌變形造成的受力變化，甚至是量測單一線軌直線度、雙線軌相對不平行度、螺帽受力以及計算軸承受力與壽命。結果顯示，單線軌直度量測實驗中智能滑塊與數位千分錶所量測之直度分別小於50Nt與4μm，而雙線軌相對平行度量測實驗中不平行度6條、5條、2條以及1條之智能滑塊與數位千分錶量測數據之間的相關係數皆大於0.5。工具機作動過程是以馬達驅動螺桿帶動工作平台於線軌上進行軸向運動，因此本研究運用相對偏移量的概念發展單一線軌直線度與雙線軌相對不平行度，以及應用靜力學取得螺帽受力與軸承受力，進而獲得軸承工作壽命，此檢測發展系統分別針對工作平台組裝前後進行量測，當無法利用馬達編碼器取得平台位置訊號時，可透過智能滑塊中的位置感測器偵測平台於線軌上的位置。實驗架構為使用數位千分錶量測線軌狀況，利用力量感測器量測線軌變形造成的受力變化，以及使用3D印表機印製的夾具夾持位置感測器。本研究發展的檢測系統可幫助調整及組裝線軌，甚至是工具機系統即時監測導軌平行度、螺帽負載變化、及軸承壽命。
A cost-effective sensor-embedded smart linear guideway type (LGT) recirculating linear ball bearing, or simply called carriage hereafter, is proposed. Two kinds of sensors are installed on this smart carriage. The first one is an optical mouse sensor which provides the positioning information. The second one is a tactile load cell, called film type pressure sensor, also known as Force Sensitive Resistor (FSR), which measures the forces from the guideway. These two sensors are utilized to monitor the position of the carriage on the linear guideway and the loading change caused by the deformation of linear guideway. By tracking the force variation sensed by the FSR along the guide way, the linear guideway straightness and parallelism between the two linear guideways can be determined indirectly. Results show that the differences in measuring the straightness of reference guideway between the proposed this smart carriage system and dial gauge are less than 50Nt and 4μm, respectively. Results also show that the correlation coefficients in measuring the parallelism of the two linear guideways are more than 0.5. Moreover, the force acting on the screw nut as well as the supported bearing at two ends can be determined when the worktable is traveling, which enables us to calculate the remaining useful life (RUL) of nut and bearings. A ball screw feed drive system with adjustable linear guideways was built to assess the performance of the proposed smart carriage both in its positioning accuracy and the on-line measurement of linear guideway straightness and parallelism as compared to those measured by the optical ruler and dial gauge, respectively
並聯式三軸力量感測器;Design of a parallel three-axis force sensor
No full text[[abstract]]本研究提出一個新型三維力量感測器，並分析推導感測器之轉換函數，最後再經由實驗驗證三軸力量感測器的結構和操作特性。目前雖然有許多方法可量測多軸力量，但多僅適用於微力量感測，如機械手臂觸覺力量感測，若欲設計可應用於人體運動或建築物承載等量測等，此類較大的三維（3D）力和負載之感測器，仍然是一個重大的挑戰。 本研究基於 Delta 平台架構並結合渦電流位移感測器，設計此具有較大負載之三軸力量感測器，其感測器設計變量為彈性樑主體的幾何尺寸，將採用有限元法對感測器彈性體進行數值分析，再經由三個高精度渦電流位移感測器檢測。因三軸力量所產生的微小變形，並實作此設計之原型機以進行校準測試，最後則進行性能之量測，以確認此設計之可行性與分析證確性。根據驗證實驗所記錄的數據，分析此三軸力量感測器之非線性，重複性和遲滯性等性能，此三軸力量測器皆具有優於市面現有產品之良好結果。
A novel three-dimensional (3D) force sensor is proposed in this study. The translation function applied in measurement has been derived, and the structure and performance are confirmed via experiments. In presently, many measurement techniques can be applied to measure the multi-dimensional force or torque. However, the force range of the measurement capability often limits in the micro-force, and the application is like the touch force of the manipulator figures. If we will apply the current measurement technical in large-scale forces, such as the human motion or building, these are incompetent. It is the challenge to design a force sensor with the large-scale force capability. In this study, we design a 3D force sensor with large-scale force capability based on the Delta structure with elasticity beam, and it adopts three eddy current sensors to estimate the deformation of the elasticity beam caused by the exerted force. The finite element analysis is applied to consider the design variables of the structure. Finally, the performance of the 3D force sensor, included linearity, repeatability, and hysteresis have been confirmed that is better than the current multi-dimensional force accordance with the experiment results
統計模型應用於單一牙冠自動設計;Statistical shape model applied to automatic single-crown reconstruction.
No full text[[abstract]]本研究發展出一套自動化單一牙冠快速設計系統，設計過程不僅無須手動且繁複地調整，更不需要主觀挑選牙冠模型的基底，計算速度快，三十秒內即可於缺牙處快速重建出符合牙齒形態、咬合關係與患者齒列關係的客製化牙冠。且此系統可與電腦輔助設計(Computer-aided Design, CAD)網格處理技術結合，提供較佳的牙冠形態做為初始狀態，再由CAD網格處理技術進行細部編修，將能大幅減少牙冠設計的程序與時間。本研究將統計模型方法(Statistical Shape Model, SSM)應用於自動化單一牙冠重建，以目標牙與其兩側鄰牙和對咬牙視為一個完整的個體並做為樣本，將大量的樣本集合成資料集合，藉由連貫點群飄移(Coherent Point Drift, CPD)建立樣本間的對應關係，待各樣本具備對應關係後，以主成分分析(Principal Component Analysis, PCA)分析資料集合並建構牙冠資料庫。最後將患者缺牙齒列的兩側鄰牙與對咬牙作為已知部分的目標模型，透過統計模型方法與牙冠資料庫比對已知部分目標，即可重建出符合牙齒形態、咬合關係與齒列特徵之牙冠。
In this research, we develop a novel system for fast and automatic design of a restorative crown. We name this system DSSM (Dental Statistical Shape Model). Unlike traditional crown design, there is no need to tediously and manually adjust the crown model during the process. Moreover, in traditional approach, choosing the base model is very subjective and redundant. DSSM can reconstruct the customized crown that satisfies the morphology, occlusion, and features of teeth arrangement within 30 seconds. Additionally, DSSM is fully compatible with current dental computer-aided design (CAD) technology. By providing a better geometric shape as the initial state and then fine-tuned with more CAD functions, we can greatly reduce the procedure and time for producing a dental crown.In this research, we apply statistical shape model (SSM) to automatic single-crown reconstruction. We first bring the tooth in the desired position, together with its neighbor and opposing teeth, and then take them as a complete sample set. Prior to this, a great amount of samples has already been used to compose the big data set. The correspondence between samples can be built through coherent point drift (CPD). After that, we take advantage of Principal Component Analysis (PCA) to analyze the data and construct the data library. For the dentition of patient with missing tooth, we are able to take the adjacent and opposing teeth to the desired position as the target model. Combining SSM method and the constructed data library, we can reconstruct the dental crown that is consistent with the dental morphology, occlusion, and features of dentition