| DC 欄位 |
值 |
語言 |
| DC.contributor | 電機工程學系 | zh_TW |
| DC.creator | 嚴柔安 | zh_TW |
| DC.creator | Jou-An Yen | en_US |
| dc.date.accessioned | 2016-8-24T07:39:07Z | |
| dc.date.available | 2016-8-24T07:39:07Z | |
| dc.date.issued | 2016 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=103521070 | |
| dc.contributor.department | 電機工程學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 電力系統中，電壓變動劇烈時，會導致日光燈等日常燈具光線閃變，使人眼視覺不適，並導致電力電子儀器的損害，造成電力系統中電壓變動劇烈的原因包括電弧爐、軋鋼馬達設備、高週波感應爐等非線性負載。電弧爐廣泛的應用於煉鋼工業，有低電壓和高電流的特性，藉由電弧產生高溫，用來熔解冶煉的廢料，冶煉過程中會造成電壓劇烈變動，使電壓閃爍情況嚴重，造成不可忽視的電力品質汙染問題，因此我們希望制定一個準確的電弧爐模型，改善現代電力系統中之電力品質汙染問題。
類神經網路具有強大的學習能力與解決高度非線性問題的能力，而電弧爐就是一個高度非線性負載，因此本文以類神經網路為基礎，建立電弧爐模型。本論文提出以小波轉換(DWT)與徑向基底函數類神經網路(RBFNN)為基礎，模擬交流電弧爐的動態電壓-電流特性。在模擬案例中，先以小波轉換分類資料群，再以徑向基類神經網路建構模型，並提出決定RBFNN初始值得方法，再以查找表(LUT)建立不同運轉時期電弧爐的電壓-電流特性。透過實驗得到的結果與實際量測數據相比，發現本文所提出方法可以準確的預測交流電弧爐的動態電壓-電流特性曲線。最後，根據所本文建立之電弧爐負載模型，透過 MATLAB進行完整的鋼鐵廠電力系統模擬。 本文所提出的方法也可以應用在其他高度非線性負載，評估抑制電力系統擾動裝置的影響。
關鍵字:電弧爐、輻狀基底類神經網路、離散小波轉換、動態電壓-電流特性曲線、電壓閃爍 | zh_TW |
| dc.description.abstract | When the voltage fluctuation occurs in the power systems, the lighting equipment would be disturbed to cause annoying variations may cause annoying variations in the output. In addition, the devices with power electronic would also be damaged.
The main causes of the voltage flicker are from those nonlinear loads such as electric arc furnace, motor drives in rolling mills, and high-frequency induction furnaces, etc. The device mentioned above like EAF is widely used in industry which has the characteristics of low voltage and high current to generate the high temperature to melt the materials. This melting process will cause the power quality(PQ) problems like voltage fluctuations which cannot be ignored. As the result, it is necessary to establish an accurate model of the electric arc furnace to improve the power quality of system.
It is known that artificial neural network is a powerful scheme for function learning and modeling nonlinear loads. This thesis proposed a discrete wavelet transform(DWT) and radial basis function neural network(RBFNN) based method for modeling the dynamic voltage-current characteristics of the electric arc furnace. In this study, a combination of the DWT and the RBFNN with parameters initialization algorithm is proposed to build the EAF voltage-current characteristics with enhanced look-up table for different operation stages. It is found that the estimated errors between experiment results obtained by this proposed model and measured data can be effectively reduced. Finally, the proposed EAF model would be realized with MATLAB program to verify the PQ analysis in the power system.
Keywords: Electric arc furnace, RBFNN, DWT, voltage-current characteristics, voltage fluctuations. | en_US |
| DC.subject | 電弧爐 | zh_TW |
| DC.subject | 輻狀基底類神經網路 | zh_TW |
| DC.subject | 離散小波轉換 | zh_TW |
| DC.subject | 動態電壓-電流特性曲線 | zh_TW |
| DC.subject | 電壓閃爍 | zh_TW |
| DC.subject | Electric arc furnace | en_US |
| DC.subject | RBFNN | en_US |
| DC.subject | DWT | en_US |
| DC.subject | voltage-current characteristics | en_US |
| DC.subject | voltage fluctuations | en_US |
| DC.title | 以類神經網路為基礎之時頻域混合交流電弧爐模型於電力品質分析之應用 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | A Neural-Network-Based AC EAF Model on Time and Frequency-Domain for Power-Quality Study | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |