摘要
论述了影响煤的可磨性的相关因素,重点探讨了在MATLAB环境下,利用几种不同的人工神经网络算法预测煤的可磨性指数,并将预测结果与实验测定值进行对比分析。分析结果表明,BP人工神经网络算法的预测精度明显高于广义回归网络和RBF网络,而且与实验值比较接近,因此是一种实用的估算K_(HCI)的方法。
The work presented here reports relative factors influencing coal grindability and focuses on predicting coal grindability index by several different artificial neural network eaculation methods. The comparison of the forecast result and practice shows that the foecast precision of BP network method is higher than generalized regression network, and RBF network and its result fits the experimental values well, so it's a good way to predict K_(HGI).
出处
《煤炭技术》
CAS
2003年第9期91-93,共3页
Coal Technology
