Application Comparison of Neural Network and Adaptive Neural Fuzzy Inference System to the Forecasting of Flue gas quality of incineration plants

• Main Authors: Chien-Ku Chen, 陳建谷
• Other Authors: Terng-Jou Wan
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/80987321869455817201

碩士 === 國立雲林科技大學 === 環境與安全工程系碩士班 === 91 === In Taiwan, the existing landfill sites reach saturation population density arising and human activities. The thermal treatment process of solid waste will be the most an important method in the future. Because of the secondary pollution generation by combustion thermal process will result in the pollution problem of ambient environment. Specially, the emission of the flue gas will influence ambient gas quality and human health. Gas pollution emission control becomes a major consideration in the design and operation of incineration plants. In recent years, fuzzy inference system and neural-network play important roles for artificial intelligence of forecaster. In this study, By Pearson and Grey Relational analyzing the data from large-scale urban solid waste incineration plants, this research will discuss the influence between operation condition and the density of the released flue gas and selected representative parameters . Back-propagation neural network、Fuzzy Inference System and Adaptive Neural Fuzzy Inference System application for flue gas prediction of solid waste incineration plants. The flue gas monitoring items include O2、HCl、NOx、SO2、CO、HF、CnHm and opacity. This study major predict NOx and build Predictive Emissions Monitoring System to accurate prediction at flue gas pollution emission concentration. Furthermore prediction the concentration of the flue gas through the change of operation conditions to reduce gas pollution treatment and operation cost.