基于模糊聚类神经网络的煤炭资源等级划分方法——以内蒙古煤炭资源预测区为例  

The Classification of Coal Resources Based on Fuzzy Clustering Neural Network ——Taking Coal Resource Prediction of Inner Mongolia as an Example

作　　者：严群[1,2] 孙杰 王秋生[3] 王庆峰 田力 

机构地区：[1]中国矿业大学(北京)地球科学与测绘工程学院,北京100083 [2]中国煤炭地质总煤炭资源信息中心,北京100070 [3]河北煤田地质局第二地质队,河北邢台054000

出　　处：《中国煤炭地质》2014年第5期30-33,57,共5页Coal Geology of China

基　　金：中国地质调查局地质调查工作项目(12120113092400)

摘　　要：依据第三次煤田预测评价指标体系,选取影响煤炭资源自然禀赋的9个指标,针对内蒙古煤炭资源预测区的数据特点,设定指标标准值和加权值。应用模糊聚类神经网络模型,对标准值和10个内蒙古煤炭资源预测区样本进行规格化处理,找出聚类中心,通过计算样本与聚类中心的欧式距离,得出样本所属等级。划分结果:呼伦湖预测区煤炭资源为一等资源,鼎辉、东胜煤田4预测区等为二等资源,得尔布尔外围、福山、炭井沟预测区虽为长焰煤和无烟煤,但资源规模太小,丰度低,煤层薄,为三等资源。Based on the index system of the third coalfield prediction, selecting 9 indexes of effecting coal resource natural endow-ments, set up standard and weight value according to the characters of the coal resource predicted area, Inner Mongolia. Applying with the model of fuzzy clustering neural network make regularization on all standard values and sample values from the 10 areas of coal re-source prediction in Inner Mongolia. Pick up the cluster centers, then classify the samples through calculate Euclidean Distance be-tween samples and the cluster centers. The classification result：predicted area of Hulunbuir Lake is the first class, predicted areas of Dinghui and Dongsheng coalfield belong to the second class, predicted areas of Deerbueroutside, Fushan, Tanjinggou belong to the third class as the small scope, low abundance, and thin layer with long frame coal and anthracite.

关 键 词：模糊聚类 神经网络 煤炭资源 等级划分 

分 类 号：O29[理学—应用数学] P628[理学—数学]