自组织竞争神经网络在决策分析中的应用

在油气田的勘探过程中,通常有多种方法可以采用,每种方法所需要的经济成本均不同,因此需要利用数学工具为油气开采选择一种最为经济的方法。这里充分利用了人工神经网络的高度智能性,给出了网络训练权值及网络训练以后的权值拓扑结构,并从理论上保证了其分类的可靠性。结合该工程中具体的数据进行了分类处理,实现了三维数据下的聚类分析,得到了良好的分类效果。最好对该算法进行了总结和分析,表明该算法具有很大的灵活性和可扩展性。

TIAN Qing-jun THE APPLICATION OF SELF-ORGANIZING COMPETING NEURAL NETWORK IN DECISION MAKING ANALYSIS

To explorate oil and gas resources, there are many techniques to be able to be adopted. Different technique has different economic cost, so that we take advantage of mathmatic tools to make decision to adopt the most economic technique for the exploration of oil and gas. The paper utilizes high aptitude of ANN and gives its network training weight and the topologic structure of the network after training. In the meantime, it pledges its reliability of classification in the theory. Using data of engineering, it complements clustering analysis under 3D condition and obtains better results of classification. The results also show that the method has better agility and extension.