摘要
利用人工神经网络从已有的炉渣分析数据中通过训练求得炉渣中磷含量与相关因素之间的非线性关系,从而预测炉渣中磷的含量。采用的神经网络结构为3-8-1的形式,学习算法采用BP算法。结果表明,网络模型有较高的预测精度,可用于炉渣磷含量的预测和控制。
This paper introduces artificial neural network method to study the non-liner relationship between phosphorus content of bira nest and its correlative factor content from the analysis data.The phosphorus content of bira nest can be predicted by trained neural network. In the study,a neural network with 3-8-1 structure was used.The learning algorithm was BP(back-propagation) algorithm.The result showed that the precision of predicting was higher and this method could be used to predict and control the phosphorus content of bira nest.
出处
《广东化工》
CAS
2005年第5期39-41,共3页
Guangdong Chemical Industry
