基于灰色系统理论的瓦斯含量多变量可视化预测方法实现

摘要

针对煤矿瓦斯含量预测采用单因素梯度法存在可靠性不高的问题,提出了一种基于灰色系统理论的瓦斯含量多变量可视化预测方法,建立了预测模型的总体框架,给出了动态链接库实现灰建模数值算法的流程；并结合某矿实际,详细介绍了该预测方法的具体实现过程:首先通过数字化瓦斯地质图获取瓦斯含量原始数据,设置建模所需的相关参数,建立瓦斯含量多变量预测模型,选择瓦斯含量点击预测或批量预测,对预测结果的误差进行分析并与图形信息结合起来显示.该预测方法提高了预测的效率及决策的科学性,为煤矿日常瓦斯含量预测、管理及决策提供了一个直观、方便、高效的可视化预测手段.%In view of problem of low reliability caused by single-factor gradient method for predicting gas content in coal mine,the paper proposed a method of visualization multivariable prediction for gas content based on grey system theory,and established overall framework of the prediction model,presented flow of grey modeling numerical algorithm realized by dynamic link database.It detailedly introduced realization process of the prediction method combined with real condition of a Mine as follows:obtaining original data of gas content from digital gas geologic map,setting up related parameters for modeling,establishing multivariable prediction model for gas content,choosing point or batch prediction for gas content prediction,analyzing error of prediction result and displaying the result with graphic information.The prediction method improves prediction efficiency and scientificity of decision making,provides an intuitive,convenient and efficient visualization prediction method for daily gas content prediction,management and decision making in coal mine.