Chung Cheng University Institutional Repository
Not a member yet
889 research outputs found
Sort by
使用低功耗振盪器讀出電路之高靈敏度CMOS MEMS壓力感測系統;Design of a Highly Sensitive CMOS MEMS Pressure Sensing System with Low-Power Oscillator-Based Readout Circuits
No full text[[abstract]]本論文實現一個以CMOS MEMS製作的電容式與壓阻式壓力感測器系統，將兩種感測器整合為單一結構，以PDMS包覆感測器結構，電容陣列結構作為壓阻感測器薄膜覆蓋區域，透過電容施壓牽動邊緣的壓阻感測器，使感測電容與壓阻可以同時變化，藉此提高感測器的靈敏度。此壓力感測系統將感測器與感測電路完全整合於同一晶片，以標準CMOS製程TSMC 0.35?m 2P4M製作完成，晶片面積為2.5?mm?^2。感測元件透過乾蝕刻與濕蝕刻完成感測器的整合，感測電路採用弛張振盪器為基礎的週期調變電容數位轉換器，弛張振盪器將感測訊號轉換為正比於頻率的輸出訊號，利用弛張振盪器本身低功耗的特性，降低整體電路的功率消耗。電容式與壓阻式壓力感測器透過弛張振盪器RC時間常數與時間長度成正比的關係，當電容與壓阻同時改變時，得到更明顯的RC時間變化。感測弛張震盪器頻率操作在1.02MHz，靈敏度為200Hz/fF，整個晶片的功率消耗為122uW。
The thesis presented a capacitive and piezoresistive pressure sensor system fabricated by CMOS MEMS method. The capacitive and piezoresistive sensor are integrated in a single structure. The capacitive array is used as piezoresistive sensor film, which cover by PDMS. By operating two sensors together, sensitivity can be increased. The pressure sensor and the sensing circuit is integrated in a single chip by standard CMOS process of TSMC 0.35μm 2P4M and the area of whole chip is 2.5?mm?^2. The sensor can be fabricated by post process of dry etching and wet etching method. The sensing circuit adopted relaxation oscillator followed by period modulation capacitance-to-digital converter. The relaxation oscillator transfers sensing signal proportional to frequency output signal. The total power is reduced through the low power characteristic of relaxation oscillator. The more obvious RC time constant changes can obtained when capacitive and piezoresistive changes together. As a result, the relaxation oscillator operating in 1.02MHz and the sensitivity of the oscillator is 200Hz/fF. The total power of the whole chip is 122uW
應用無線天文儀表考量增益壓縮設計之E-Band混頻器;E-Band Mixer with IP1dB Design Consideration for Radio Astronomical Instrumentation
No full text[[abstract]]此論文主要的內容是應用在 此論文主要的內容是應用在 E-Band天文望遠鏡之高 天文望遠鏡之高 增益壓縮 (IP1dB)的降 頻混器， 他的主要特色有兩個： 他的主要特色有兩個： 第一個特色 是其優異的轉換 增益 以及其輸入 以及其輸入 輸出的頻寬，此混器在 本振 端埠使用頻寬為 79~94 GHz，並在本振功率 輸入 9.45 dBm時，可以將 射頻端埠 67~90 GHz的輸入 訊號，降頻為 4~12 GHz的低 頻 IF輸出 訊號，而且可以將混頻器的轉換 訊號，而且可以將混頻器的轉換 增益 保持在 -8 dB；另一個特色是這 ；另一個特色是這 個混頻器 的輸入增益壓縮點 ，它的 IP1dB大小至少超過 大小至少超過 1 dBm，推斷此混頻器 的 IP1dB最高可達 9 dBm。此降頻混器在電晶體的 此降頻混器在電晶體的 此降頻混器在電晶體的 選擇 ，使用 的是 Cascode架構 ；製程方面則是使用了穩懋 0.1-μm 砷化鎵 pHEMT；電路傳輸線設計方面 ， 使用了共面波導傳輸線 (GCPW)。整體混頻器電路的大小為 2 μm * 1.4 μm，混 頻器完成後，將可應用在 ALMA計畫的天文收發系統中。
In this paper, we presents an E-band high IP1dB mixer for astronomical telescope application. This device has two features, one is its great conversion gain and broad I/O bandwidth. For this work, the available bandwidth of LO, RF and IF port is 79-94 GHz, 67-90 GHz and 4-12 GHz, respectively. With LO input power of 9.45 dBm, this mixer can down-convert the E-band signal from 67-90 GHz to 4-12 GHz, with conversion gain of -8 dB. The other excellent feature is IP1dB. This E-band mixer has high IP1dB of more than 1 dBm. A speculation can be concluded that the IP1dB is estimated about 9 dBm. This work use the cascode topology for the transistors. This work is implemented in WIN 0.1-μm GaAs pHEMT process and grounded coplanar-waveguide (GCPW) is used for the circuit design. The overall mixer size is 2 μm * 1.4 μm. After the mixer circuit is completed, it could be applied in the astronomical telescope receive system in ALMA project
台電高佔比再生能源併入下輸電系統壅塞分析與視覺化介面開發;Visualization-based congestion analyses for Taiwan’s transmission system after integrating a large amount of renewable energy
No full text[[abstract]]因應全球再生能源發電比例日益提高，電網的結構變化和傳輸網路的壅塞在許多國家中受到相當的重視。我國政府提出的能源政策為2025年離岸風電裝置容量達到3GW，與傳統發電廠相比，再生能源出力具有隨機波動的特徵，勢必對電網造成一定的衝擊，其中一個主要的影響是造成傳輸網路的線路流量過載。因此，在高占比再生能源併網前後了解系統目前的傳輸瓶頸是重要的。本文使用台灣電力系統參數進行模擬不同季節、不同時段下既有的電力系統傳輸線路瓶頸，以及未來在台灣西部沿海變電所併入離岸風力發電與西部各縣市併入太陽能發電的情境分析，包含可能壅塞線路的評估。為避免線路壅塞導致再生能源無法傳輸至用戶，若能即時處理未來大量再生能源併入電網後的線路壅塞，必能提高能源使用效率，並建立一個友善的併網環境，將其模擬結果提供給系統規劃參考。此外，本研究初步構構一個台電輸電系統視覺化介面，藉此可清楚地呈現輸電線路壅塞瓶頸，供系統規劃與調度人員參考。
With the increase of installed capacity of renewable energy, the transmission expansion and line congestions play the vital role in many countries. According to the data from Industrial Technology Research Institute (ITRI), in 2025, the target of total capacity for offshore wind farm in Taiwan will be around 3 GW. Compared to conventional power plants, renewable power output has random and fluctuant characteristics. Additionally, the existing power grids would have insufficient capacity to transfer those additional renewable energies, which increases the risk of the overload on the transmission lines. Transmission congestions would increase system operation cost and waste nature resources. Therefore, it is significant to identify transmission bottlenecks before and after a large scale of renewable energy integration in advance. In this thesis, the data of generation and load recorded by Taipower was utilized to investigate the possible transmission congestions. Moreover, various scenarios analyses by considering different renewable energy integrations were implemented. The result of this thesis can provide a reference which can help system operators to establish a friendly environment for renewable energy integration and to avoid high transmission cost caused by line congestions. Additionally, this thesis developed a visulization interface for the Taiwan transmission system. It can present the congestion bottlenecks clearly and provide a reference to system planners and operators
結合頻譜分析與機器學習方法檢測配電系統之故障位置;A Hybrid Approach for Locating Fault in Distribution System by Measured Frequency Spectra and Learning Algorithm
No full text[[abstract]]Locating the fault in a distribution system is still a challenging reliability issue. It is due to not only the complicated topology such as unbalanced system, DG integration, shunt and series compensation, and laterals but also some fault location parameters such as load variation, fault inception angle (FIA), fault resistance and the presence of DG integration. This thesis proposes an algorithm in which fault location is found by using the hybrid approach of signal analysis approach and learning based approach. The transient signal recorded at the substation will be transferred to the frequency domain by Fourier transform. The frequency spectrum at each location is different due to the system configuration. Support vector machine classifier and regression analysis are used to recognize the fault area and distance, respectively. Particle swarm optimization is used to optimize the parameters of the learning algorithm. Time series forecasting technique ARIMAX is used to forecast short-term load profile in order to select the appropriate model. The simulations have modeled in ATP/EMTP according to IEEE 34-bus, 123-bus benchmark test feeder and Taipower’s distribution system. Test results show that frequency spectrum can overcome most aforementioned obstacles with high accuracy for detecting fault location.
Locating the fault in a distribution system is still a challenging reliability issue. It is due to not only the complicated topology such as unbalanced system, DG integration, shunt and series compensation, and laterals but also some fault location parameters such as load variation, fault inception angle (FIA), fault resistance and the presence of DG integration. This thesis proposes an algorithm in which fault location is found by using the hybrid approach of signal analysis approach and learning based approach. The transient signal recorded at the substation will be transferred to the frequency domain by Fourier transform. The frequency spectrum at each location is different due to the system configuration. Support vector machine classifier and regression analysis are used to recognize the fault area and distance, respectively. Particle swarm optimization is used to optimize the parameters of the learning algorithm. Time series forecasting technique ARIMAX is used to forecast short-term load profile in order to select the appropriate model. The simulations have modeled in ATP/EMTP according to IEEE 34-bus, 123-bus benchmark test feeder and Taipower’s distribution system. Test results show that frequency spectrum can overcome most aforementioned obstacles with high accuracy for detecting fault location
應用區間二型T-S模糊輸出回授追蹤控制系統之單相雙向換流器研製;Design and Implementation of a Single-Phase Bidirectional Inverter Using IT2 T-S Fuzzy Output-Feedback Tracking Control Systems
No full text[[abstract]]本文設計區間二型Interval Type 2 (IT2) T-S (Takagi-Sugeno)模糊輸出回授追蹤控制系統，並應用於2kW單相雙向換流器之研製，使電路可操作於市電併聯及整流模式。首先，考慮功率開關導通壓降、二極體導通壓降及被動元件內阻之電路，推導換流器操作於市電併聯及整流模式之狀態空間模型。接著將前鑑部參數不確定因素之變動納入，建立IT2 T-S模糊模型。然後，合併IT2 T-S模糊輸出回授追蹤控制器，形成換流器閉迴路IT2 T-S模糊輸出回授追蹤控制系統。本文第二章介紹雙向換流器，以換流器操作於市電併聯及整流模式之實際電路圖說明電流流動方向，並推導正、負半週狀態空間方程式；第三章為本論文重點，設計IT2 TS模糊輸出回授追蹤控制系統。本文提出兩個IT2 T-S模糊輸出回授追蹤穩定定理，定理一為推導考慮外在干擾且滿足 性能指標之LMI(Linear Matrix Inequality)穩定條件式；定理二為同時考慮外在干擾與實際電路之模式不確定性強健LMI穩定條件。最後，以1kW、1.5kW、2kW三組功率之電腦模擬與實驗，完成IT2 T-S模糊輸出回授追蹤控制系統應用於單相雙向換流器之電路分析與驗證。經由抗干擾、模式不確定性之電腦模擬分析與實驗結果，均顯示本文設計之IT2 T-S模糊輸出回授追蹤控制器，不論在市電併聯或整流模式，均比分切合整控制法優異。
In this thesis, Interval Type-2(IT2) T-S(Takagi-Sugeno) Fuzzy Output-Feedback Tracking Control Systems is designed and applied to a single-phase bidirectional inverter which can operated in grid-connection mode and rectification mode in 2-kilo watt. First, we have to derive the state space model of the inverter which operates in grid-connection mode and rectification mode and consider voltage drop of power switches, diode and passive components. After that, merging IT2 T-S fuzzy output-feedback tracking model and tracking to form IT2 T-S fuzzy output-feedback tracking closed loop control systems of the inverter. In this thesis, Chapter 2 will introduce bidirectional inverter. We will describe direction of current of inverter which operated in grid-connection mode and rectification mode and derive state space equation in half and negative half week. Chapter 3, the focuses in this article, proposes design of IT2 T-S output-feedback tracking control system. This thesis proposes two IT2 T-S fuzzy output-feedback tracking stability theorems. Theorem 1 is that deriving LMI(Linear Matrix Inequality) stability conditions which consider external disturbance and satisfy H infinity performance index. Theorem 2 is the robust LMI stability conditions which consider external disturbance and model uncertainty of real circuit simultaneously. Last, we have experiments to analyze and prove the performance of IT2 T-S fuzzy output-feedback tracking control system which is applied to the single-phase bidirectional inverter in 1 kilo, 2 kilo and 3 kilo watt. Then we have anti-disturbance and model uncertainty experiments to prove the method proposed in this paper which has better performance than D-? method whether in grid-connection mode or rectification mode
基於時變慣性響應與下垂控制之風場頻率調節研究;Study of Frequency Regulation by a Wind Farm Using Time-Varying Inertia and Droop Controls
No full text[[abstract]]隨著再生能源的興起，輸電系統操作者對於電力系統頻率穩定度變得更加的重視，尤其是在小型的電力系統中，傳統的風力發電系統因為與電力系統頻率之間呈現解偶關係使得無法提供電網適時調節系統頻率的能力，因此當傳統的火力發電機組漸漸的被風力發電機取代，風力發電系統對電網頻率調節的議題就變得越來越重要。 近年來很多的相關文獻都致力於研究有關風力發電機的慣性響應和下垂控制對於電力系統頻率調節的能力，為了實現此種調頻能力，我們必須在電力電子轉換器上加上慣性響應變數和下垂控制變數的雙迴路控制策略去決定輸入到轉子側變換器的實功參考值訊號，慣性響應運用在快速且暫態的頻率調節，而下垂控制則提供了較長且穩態的頻率調節，本篇論文基於目前文獻中所發展的風力機調頻控制模型提出結合了在不同風速和時間下的慣性響應變數和下垂控制變數之新的控制策略，此種新的控制策略整合了慣性響應控制、轉子速度控制和槳距角控制進而發展出一套完整的頻率調節控制機制，有別於以往的風力發電機頻率調節架構，新的控制策略在二匯流排測試電力系統和實際島嶼電力系統模擬中驗證，結果顯示此篇論文所提出新的控制方法對於電網頻率調節能力均有相當大的改善。
With the rising of renewable energy, frequency stability problem has caused a particular concern for the transmission system operators (TSOs) especially in a small power system. Traditional wind generation system does not provide the appropriate capability of frequency regulation because of the decoupling from the power grid. Therefore, as conventional thermal generators are replaced partly by wind generators, the issue of wind generation system about frequency regulation has become more and more important. Several literatures have studied on the inertia control and droop control of wind generators to support frequency regulation capability in the power grid. To develop the functionality of frequency regulation capability, it is necessary to add control loops with inertia and droop variables to determine the power reference that is an input signal for rotor side converter (RSC) of DFIG. The inertia variable represents the transient frequency response and the droop variable affects the steady state frequency response. This study proposes a new control strategy for the inertia and droop variables to implement coordinate control in different wind speeds and time based on the current frequency regulation model in recent literatures. The new control strategy integrates the inertia control, rotor speed control and pitch angle control to accomplish the complete frequency regulation mechanism. Compared to the frequency regulation framework of the past, the new proposed method is verified in a two-bus test power system and an island power system. Consequently, it superiorly improves the performance of frequency regulation in the power grid
微波功率檢測器與相位不平衡檢測電路之設計;Design of Microwave Power Detector and Phase Imbalance Detection Circuitry
No full text[[abstract]]此論文為振幅與相位不平衡檢測電路設計，包含功率檢測電路的設計與模擬量測，使用內建固定相移器用於相位偵測的新概念。在功率檢測電路設計上，使用負載改良的共閘極功率檢測單元，分別針對Ku-band(12 GHz ~ 18 GHz)與超寬頻(0 GHz ~ 30 GHz)設計功率檢測電路，使用1P6M 0.18 um CMOS 製程；與V-band功率檢測電路使用1P9M 90-nm CMOS。在相位檢測上，由於當兩路訊號以較低相位差進行振幅疊加時，其每度相位差經由功率檢測器之輸出電壓變化低，較難以輸出電壓判別相位，因此在訊號合成器加上內建90度相移機制，改變兩訊號合成時相位，改良低相位差區域相位判別度。實作上設計3.5 GHz相關電路，分別需要2路相位可調訊號源、內建90度相移之訊號合成器與功率檢測器，並進行模擬與量測比對與初步演算法的設計。
This research topic is the microwave power amplitude and phase imbalance detection. The study also presents the design and implementation of power detector using load-improved common gate power detection unit, and proposes the new concept of phase imbalance detection by using signal combiner with embedded phase shifter.Two of fabricated power detectors, Ku-band power detector, ultra-wideband power detector are design by 0.18 um CMOS process, and the V-band power detector is by 90 nm CMOS process. All of them are introduced in chapter III.When using the superposition of signals for phase detection, it may be hard to discriminate the phase difference between the signals as their phase difference is small. In this study, the way employing an embedded phase shifter added into both path of signal combiner is proposed, it changes the relative phase when two signals superimpose, improving the discrimination of phase difference for the low phase difference region. This phase detection system is implemented at 3.5 GHz, and composed of four components, two signal sources with 0?~180? tunable phase, signal combiner with embedded phase shifter, power detector and initial design of algorithm
基於腦波的疲勞駕駛偵測系統;An EEG-based Driver Drowsiness Detection System Design
No full text[[abstract]] 每年都有數以萬計的車禍發生，其中有大部分的車禍是因為疲勞所引起，疲勞駕駛變成相當嚴重的問題，最糟的情況還有可能成為致命的車禍，所以我們必須使用有效的方法來偵測疲勞駕駛才能避免嚴重的交通意外事故發生，因此我們希望藉由提出一個創新的系統來解決這一個嚴重的問題。 為了偵測疲勞駕駛，我們提出了一個基於可穿戴式單通道腦波儀實作的疲勞偵測系統，這個單通道腦波儀並不昂貴且不會侵入人體，因此駕駛可以負擔儀器的金額，而且穿戴起來並不會不舒服，不會影響對車輛的操作，駕駛在穿戴腦波儀的同時便向運算平台傳送腦波資料，如此可減少偵測疲勞的操作時間，而且能在偵測到疲勞時將危險警訊回饋給駕駛知道。我們的系統使用兩次驗證的方式來降低誤報的可能性，如此可避免造成駕駛的困擾。 經實驗證實我們的系統能在短時間分辨駕駛的精神是清醒或是疲勞的階段。
Every year tens of thousands of traffic accidents occur. Most of them are related to fatigue. Fatigue driving has always been a very serious traffic problem. In the worst case, such traffic accidents can be fatal. In order to ensure road safety and increase driving safety, we must use an effective method to detect the driver’s drowsiness level to prevent serious traffic accidents. Thus, we expect to be able to build an innovative system to solve this serious problem.In order to detect driving fatigue, we propose a drowsiness detection system which is based on wearable single channel EEG headset device. The single channel EEG instrument is not expensive so driver can afford it and it is a non-invasive instrument that can reduce the wearing discomfort and does not affect the driver’s operation of the vehicle. The driver wore the EEG instrument at the same time transmit their EEG data to a computing platform. In this way it can effectively reduce the operation time to do drowsiness detection and the danger warning can be sent back to the driver as a feedback. Our work uses two drowsiness verification step to reduce the number of false positives.Experiments show that our system can effectively distinguish between the state of awake and drowsiness of driver in short time
使用指令位置命名記憶體配置之研究;Study On Memory Allocation Using Instruction Address
No full text[[abstract]]非揮發性記憶體（Non-Volatile Memory）是一種次世代的記憶體，其特性在於該記憶體只需使用些微的電量或幾乎不耗電就能做到資料的儲存，因此我們可以利用這種記憶體來對傳統的記憶體架構作優化，這將會是在裝置節能上的一大突破。本文探討的記憶體架構為混合式主記憶體（Hybrid Main Memory），即是一般的記憶體架構加入部分的非揮發性記憶體(主要探討PCM)，由於非揮發性記憶體在待機上幾乎是不耗電的，因此經過使用這種記憶體優化後的記憶體架構能夠有效降低裝置的耗電量，若能有效運用此記憶體架構，我們便能夠延長手持裝置的續航時間，也能夠對伺服器，桌上型電腦等裝置達到節能的效果。在實驗上本論文將探討混合式主記憶體的配置方式，根據混合式主記憶體其本身的特性與限制，認為有效使用該記憶體架構的關鍵在於對記憶體內資料使用方式的分析與分類，並於實驗中使用Linux內部的工具對程式做指令階層的分析，找出適用於此記憶體架構的資料擺放方式
結合IoTivity框架與互動式連線建立之NAT穿透技術;A NAT Traversal Technique Integrated with IoTivity Framework and Interactive Connectivity Establishment
No full text[[abstract]]隨著物聯網技術迅速地發展，人們的日常生活中已經逐漸充滿物聯網設備的影子，因此現今發展出許多不同規格的物聯網系統，導致使用者無法只使用同一套物聯網系統就能控制所有物聯網設備，有鑑於此，本論文採用IoTivity框架來建置出一套物聯網系統，使得即便是由不同家廠商所製造的物聯網設備，只要其符合IoTivity的規範就可以透過此系統來進行控制，讓使用者可以只使用這一套物聯網系統就能夠控制自己擁有的所有物聯網設備。當代的網路環境日趨複雜，因此若當由IoTivity所建置的系統位於私網環境下的話，是無法讓不是位於同一私網環境下的使用者直接連線進來使用的，故此，本論文還將此系統結合互動式連線建立技術來進行NAT穿透，讓無論身處何種類型NAT後的使用者都可以透過建立P2P連線的方式來使用此系統。
Along with the rapid development of the Internet of things technology, people's daily life has gradually filled with IoT devices. Therefore, many different specifications of IoT systems have been developed, which caused users cannot use the same IoT system to control all IoT products from different manufacturers. Therefore, this paper uses IoTivity Framework to build an IoT system that can control all IoT products which conform to IoTivity specifications, even if it is produced by different manufacturers and allow users to control all of their own IoT devices by using only this IoT system.Nowadays, the network environment is becoming more and more complex, if the system built by IoTivity is in the private network environment, then users who are not in the same private network environment cannot use it directly.As a result, this paper also combines this system with Interactive Connectivity Establishment to achieve NAT Traversal, so that users who are behind any type of NAT can use this system by establishing a P2P connection