摘要
针对目前国内粉煤灰分类的片面性，结合人工神经网络理论，提出了基于人工神经网络的粉煤灰科学分类．该方法充分考虑了粉煤灰的火山灰活性，并考虑了细度、玻璃体、烧失量、K2O、SO3和CaO多种因素对分类的影响．实例表明，按本文提出方法建立的网络模型较合理，且其精度也较高，克服了以往分类法的片面性，较全面地反映了粉煤灰的品质性能，从而为粉煤灰的多元化利用奠定基础．
For the one sidedness of fly ash classification in China, a ANN-based scientific classification of fly ash was proposed in combination with the theory of Artificial Neural Network. The method considers the pozzolanic activity of fly ash, degree of finess, vitric, loss on ignition, K20, SO3 and CaO that are vital to scientific classification. The result shows that the proposed ANN model is more reasonable and its precision is higher than the former methods. It can reflect the quality of fly ash all-roundly, and lay a foundation for the diversification utilization of fly ash.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2005年第B08期107-110,共4页
Journal of China Coal Society
关键词
火山灰活性
粉煤灰分类
多元化利用
人工神经网络
pozzolanic activity
classification of fly ash
diversification utilization
ANN
