摘要
近年来,找盲矿法从原生晕叠加理论发展到构造叠加晕理论,显示高矿区的成矿过程具有多期、多阶段叠加的特征,这增加了盲矿信息源、提高了找矿预测精准度。同时,构造叠加晕法在盲矿预测模型建立过程中存在非线性、稀疏化及结果可靠性评价等难点问题亟待解决,本文结合相关向量机的优点,对盲矿预测的构造叠加晕法进行参数改进,提出一种新的基于自适应进化相关向量机的构造叠加晕盲矿预测模型,以期提高深部盲矿预测的准确性。
In recent years,the blind mine method has been developed from the original halo superposition theory to the structural superimposed halo theory,which shows that the mineralization process in the high mining area has multi-period and multi-stage superposition characteristics,which increases the blind mine information source and improves the prospecting prediction accuracy degree.At the same time,the construction superimposed halo method is difficult to solve in the process of establishing blind mine prediction model,such as nonlinearity,sparseness and reliability evaluation.The paper combines and improves the correlation vector machine and blind mine prediction structure.Based on adaptive evolution correlation vector machine to construct a superimposed bluff mine prediction model,in order to improve the accuracy of deep blind mine prediction.
作者
于聪聪
YU Cong-cong(Jiangxi Bureau of Geology and Exploration,Nanchang 330030,China)
出处
《宜春学院学报》
2019年第3期48-52,共5页
Journal of Yichun University
关键词
盲矿
构造叠加晕
相关向量机
预测模型
准确性
blind mine
the original halo superposition theory
thestructural superimposed halo theory
adaptive evolution correlationvector machine
