基于Box Behnken设计的煤导热系数分析与预测

为研究分析水分、固定碳及密度3个因素对煤导热系数的影响，选取多个矿井不同煤质的煤作为实验煤样，使用NETZSCH LFA457型激光导热系数测试仪进行测定，并引入二次响应曲面试验设计思想，采用Design Expert软件进行煤质指标的Box Behnken试验设计，按设计表筛选实验数据，构建了导热系数多因素、多水平影响下二次回归响应曲面模型，并对不同采样地点的煤导热系数进行了分析与预测。结果表明:二次回归方程P值小于0.000 1，极显著，失拟项为0.072 4，不显著，回归方程具有统计学意义；当水分在0.9%~10.9%，固定碳在49.3%~83.6%，密度在1.2~1.92 g·cm³之间，对煤的导热系数一次项影响程度为密度>固定碳>水分；二次项影响程度为固定碳和密度>水分和密度>水分和固定碳，且固定碳和密度之间存在交互作用，水分和密度、水分和固定碳之间无交互作用；应用响应曲面模型预测的煤导热系数误差为4.3%，满足精度要求。

Analysis and prediction on thermal conductivity of coal based on Box Behnken design

In order to analyze the influence of water content, fixed carbon and density on the thermal conductivity of coal, the coal with different coal quality from several mines were selected as the experimental coal samples, then the measurement was conducted by using NETZSCH LFA457 laser thermal conductivity tester, and the quadratic response surface methodology in experiments design was introduced to carry out the Box Behnken experiments design of the coal quality index by using the Design Expert software. The experimental data were selected according to the design tables, and a quadratic regression response surface model of the thermal conductivity under the influence of multi-factors and multi-level was established, then the thermal conductivity of coal from different sampling sites were analyzed and predicted. The results showed that the P value of the quadratic regression equation was less than 0.0001 with extremely significant, and the lack of fit was 0.0724 with no significant, so the regression equation had the statistical significance. When the water content was between 0.9% and 10.9%, the fixed carbon value was between 49.3% and 83.6%, and the density was between 1.2 g·cm3 and 1.92 g·cm3, the influence degree of the first terms on the thermal conductivity of coal was in the order of density, fixed carbon and water content from large to small, and the influence degree of the quadratic terms was in the order of fixed carbon and density, water content and density, and water content and fixed carbon from large to small. Meanwhile, there existed the interaction between fixed carbon and density, but no interaction between water content and density, and water content and fixed carbon. The error of the thermal conductivity of coal predicted by the response surface model was 4.3%, which could meet the requirement of accuracy.