摘要
该文介绍了一个基于煤岩反射率分布并使用神经网络进行煤岩成分分析的方法。该方法利用神经网络在优化计算方面的强大功能,同时又考虑到煤岩分析系统的快速性和稳定性,设计了一个资源开销很小、拓扑结构比较简单的神经网络和与之相适配的快速算法来确定煤种比例,实验表明该方法的分析结果和分析速度均令人满意。该方法对于各种混合物质通过显微成像进行成分分析有一定的参考价值。
In this paper a method to perform coal analyzing based on coal reflectivity by a simple Neural Network is introduced.By using the power of Neural Networks in the field of Pattern Recognition and Computing Optimize,and considering the speed and stability which the coal-analysis-system requires,this paper constructs a Neural Network of less resource spending and rather simple topology,together with an adaptive fast-algorithm to compute the ratios of coal components.It has been proved by experiments that both the results and speed of this method meet the demanding.This method can be referenced by any components analysis system of mixture.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第35期203-205,208,共4页
Computer Engineering and Applications
基金
国家自然科学基金(编号:69975004)
关键词
煤岩
反射率分布
神经网络
快速算法
煤种比例
成分分析
Coal,Reflectivity,Neural Network,Fast-algorithm,Ratios of coal components,Components analysis
