Interior Lighting Design by Using Artificial Neural Network

• Main Authors: Chung-Yung Chiang, 江仲勇
• Other Authors: Horng-Ching Hsiao
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/47186158277284567819

碩士 === 國立臺灣科技大學 === 電機工程系 === 91 === The purpose of this thesis is to research on the application of interior lighting design by using artificial neural network, The classrooms are chosen to be the typical lighting zones of the research. In this work, back-propagation neural networks is applied to learn and generalize the “internal rule” of the classroom lighting design input parameters and lighting quality indicators. Regarding the four input parameters (luminaire type, luminaire arrangement, luminaire orientation, and luminaire spacing) affecting the classroom lighting quality, we used neural networks to predict the three lighting indicators (average illumination, visual comfort probability, and lighting uniformity), and to evaluate overall lighting quality with additional consideration of the power consumption indicator. Besides, the design input parameters were examined the probabilities which related to a low lighting quality. The experimental results show that it is applicable to back-propagation neural networks to predict lighting indicators and to evaluate quality, as neural networks has been extensively applied in the domain of prediction and diagnosis.