改进灰色Verhulst模型在巷道变形量预测中的应用

Application of improved grey Verhulst model in roadway deformation prediction

基于复杂地质条件下巷道围岩相对变形量呈S型的特点,研究构建了灰色Verhulst巷道变形量预测模型。针对预测函数构造过程中参数的取值影响预测精度的问题,从灰色Verhulst模型背景值的误差来源入手,重构背景值公式;并根据最小二乘法原理及灰色系统理论的新信息优先原理,将第二次测量的观测值作为初始条件,建立改进背景值与初始条件相结合的灰色Verhulst预测模型。经金山店铁矿破碎带巷道工程实际应用验证,改进后灰色Verhulst模型的平均绝对误差由68.832降低到2.587,预测精度由不合格提高到优,且巷道变形量的极限值为0,符合试验现场巷道稳定的实际情况。

Based on the S-type characteristics in the relative deformation of roadway surrounding rocks under the complex geological conditions,the grey Verhulst roadway deformation prediction model was researched and built.Addressing the problem of parameters valuation affecting the prediction accuracy in the prediction of functional structure,the formula of background value was reconstructed proceeding with the error source of the background value for the grey Verhulst model background value.According to the principle of least square method and the new information priority principle in the grey system theory,the observed values by second measure served as the initial conditions,the grey Verhulst model was built with the combined background value and initial conditions.The improved grey Verhulst model applied in the roadway project in the crush area of Jinshandian Iron Mine showed that the average absolute error was reduced from 68.832 to 2.587 and the prediction accuracy was increased from unqualified to desirable The roadway deformation limit value was 0,which complied with the actual situations of roadway stability on the test site.