基于GA-SVM大采深条件下底板破坏深度回归分析

Regression Analysis of Floor Damage Depth in Deep Mining Based on GA-SVM

在收集大采深条件下底板破坏数据样本基础上,充分发挥遗传-支持向量机在小样本环境下具有分类强、适应度好的特点,构建大采深条件下底板破坏深度与开采深度、煤层倾角、煤层开采厚度、工作面长度、底板抗破坏能力和断层或破碎带之间的非线性模型,采用"留一下验证法"验证该模型的泛化度,经验证该模型具有很强的泛化度。

Builds a nonlinear relationship model between the depth of damage floor and mining depth, coal seam dip angle,the thickness of coal seam mining,the length of working face,the ability of floor resisting damage and fault or fracture zone on the condition of deep mining on the basis of collecting the sample of damage floor data on the condition of deep mining and gives full play to the characteristics of strong classification and good adaptability that genetic algorithm and support vector machine has in small sample environment, validate the generalization of this model by leave-one-out method and proved this model has a strong degree of generalization.