煤炭质量是煤炭企业的生命、是占领市场的重要保证。采煤工作面煤质管理对煤炭质量有决定性作用。运用计算机及网络技术实现采煤工作面煤质预测及煤层煤质信息可视化，使得煤质工作面人员能够更加直观、准确的了解工作面煤层结构，对井下采煤工作具有重要的现实意义和应用价值。 本文首先介绍煤质管理总体业务流程，重点分析工作面煤质管理和煤质基本信息，以及传统月度煤质预测方法。在此基础上分别研究基于线性的多元回归和非线性的RBF神经网络预测方法。采用最小二乘法对多元线性回归系数估计，通过F检验和T检验验证回归方程的适用性；采用自组织学习算法计算RBF基函数中心，通过实验数据对预测结果比较和分析。 继而研究工作面曲线曲面插值。利用分段连续三次Bezier曲线插值回采工作面素描图，通过约束边界以及加入节点对曲线改进，使改进后的曲线避免层次越界；针对工作面煤层曲面构造分别研究反距离加权插值和普通克里金插值模型，并加以比较。 随后研究基于JavaScript技术的工作面煤质信息三维建模。首先介绍三维可视化数学原理，然后介绍JavaScript、Ajax、SVG和VML等实现技术，最后以工作煤层煤样数据为基础，将JavaScript脚本语言结合三维可视化技术，实现网络环境下煤层工作面三维动态绘制。 最后，基于Java语言的Eclipse开发环境下，建立基于Freemarker+WebWork+ Spring+Ibatis架构的煤质预测及煤质信息可视化系统。包含数据的整合、汇总和上报、报表处理、煤质预测、地质图件绘制等功能，从而验证本文研究成果的可行性与有效性。

The coal quality is the life of coal enterprises, is an important guarantee to capture the market. Coal quality management of coal face has a decisive role in the quality of coal.Using computer and network technology to realize coal quality forecast and coal quality information visualization of coal face, which will make the coal staff understand the coal seam structure intuitively and accurately, it has an important practical significance and application value for underground coal mining work. This paper first introduces the overall business processes of coal quality management, it’s emphasis is coal face’s quality management, basic information and traditional method of coal quality forecast.Based on this information, researching the prediction method of multiple regression and RBF neural network which based on the linear and non-linear. Using the least square method to estimate coefficient of multiple linear regression, verifying the applicability of regression equation by F-test and T-test; Using the self-organization learning algorithm to calculate the center of RBF basis function,then compare and analyze the forecast results of experimental data. Then examined the interpolation of coal face’s curve and surface.Piecewise continuous three Bezier curve interpolate working face sketch. The curve improved by the boundary and adding nodes, so that the improved method avoid corss-border between level curves; For coal seam surface structure were studied inverse distance weighted interpolation and ordinary kriging interpolation model and compare them. Subsequently, research three-demensional modeling of coal face information based on JavaScript technology. First introduced JavaScript, Ajax, SVG and VML, and then introduced the mathematical principles of three-dimensional visualization, based on the data of coal seam’s sample,combining the JavaScript language and 3D visualization technology, realizing the three-dimensional dynamic display of face;s coal quality information under the network environment. Finally, under the development of Java Eclipse, Buliding the coal prediction and coal quality information visualization system based on the framework of Freemarker+Webwork +Spring+Ibatis.This system contains data integration, aggregating and reporting, coal quality forecasting and geological maps,which verify the feasibility and effectiveness of the research in this article.