一种基于模拟退火算法岩体可爆性分级的投影寻踪回归方法

A Projection Pursuit Evaluation Method of Blastability Classification Based on Simulated Annealing Algorithm in Underground Rock

在分析影响岩体可爆性因素的基础上,选择岩体坚固性系数、声阻抗、炸药消耗量、平均裂隙距作为岩体可爆性分级评判指标,建立岩体可爆性分级的投影寻踪综合评价模型,利用遗传算法优化其投影方向,根据投影特征指标值对岩体可爆性进行分级评价,并将其应用于某地下岩体可爆性分级。评价结果表明,该评价模型可以避免主观因素对权重确定的影响,评价结果客观准确,为岩体可爆性综合判别提供了一种新的方法。

Based on the analysis of influencing factors for rock blastability,the consistent coefficient,acoustic impedance,explosive consumption,and mean fracture distanced of wall-rock are chosen as the criterion index for rock blastability classification.A projection pursuit evaluation model for rock blastability was established and its project direction was optimized by the Simulated Annealing Algorithm.Then the rock blastability can be evaluated and classified according to the index of projection.The model was applied to the classification of underground rock.The evaluation results showed that this model can avoid the impact of subjective factors on the classification.The results are accurate and true,thus providing a new proposal to comprehensively evaluate the rock blastability.