Utility boiler's combustion performance modeling based on modular RBF network
暂无分享,去创建一个
To optimize the utility boiler' s combustion process, a method for its combustion performance modeling based on modular Redial Basis Function(RBF) Neural Network is proposed in this paper. The whole modeling can be divided into two stages: first, get the mathematical model of carbon content of fly ash, exhaust flue gas temperature and their related input parameters; second, take the output of Neural Network as the input of boiler thermal efficiency calculation, and build a modular performance model of boiler combustion. This method can express the boiler combustion model in parts, the parts that can be described with mathematics be expressed with functions, other parts that can not be described with mathematics be expressed with RBF Neural Network. Data test and practical applications prove that this modeling method is efficient, has high precision and also meets the needs of boiler's running optimization.
[1] Naonori Ueda,et al. Optimal Linear Combination of Neural Networks for Improving Classification Performance , 2000, IEEE Trans. Pattern Anal. Mach. Intell..
[2] Sherif Hashem,et al. Optimal Linear Combinations of Neural Networks , 1997, Neural Networks.
[3] Galina L. Rogova,et al. Combining the results of several neural network classifiers , 1994, Neural Networks.