基于蚁群算法岩体可爆性分级的投影寻踪回归方法  

BLASTABILITY CLASSIFICATION METHOD OF PROJECTION PURSUIT EVALUATION BASED ON ANT COLONY ALGORITHM IN UNDERGROUND ROCK

作　　者：方崇[1] 成艳荣[2,3] 代志宏 张信贵[5] 

机构地区：[1]广西水利电力职业技术学院,南宁530023 [2]空军工程大学工程学院系统工程系,西安710038 [3]徐州空军学院机场工程系,徐州221000 [4]广西电力勘察设计研究院,南宁530023 [5]广西大学土木建筑工程学院,南宁530004

出　　处：《工程爆破》2010年第1期12-15,8,共5页Engineering Blasting

基　　金：中国-新加坡国际合作项目<地下岩洞爆炸损伤的模型试验研究>;广西自然科学基金(桂科自0728037)联合资助

摘　　要：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,利用蚁群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。After the factors of influencing rock blastability were analyzed, the consistent coefficient, acoustic impedance, explosive consumption, and mean fracture distanced of wallrock were chosen as the criterion indices for rock blastability classification. A rock blastability evaluation method was established and its projection direction was optimized by Ant Colony Algorithm. The value of projection characteristics was used to evaluate rock blastability and was applied to classify certain underground rock blastability. The evaluation results showed that this method avoided the subjective weighting of judgers and the evaluation results were accurate. Projection Pursuit is a new evaluation method of judging and classifying rock blastability, which provides a new way to evaluate the rock blastability.

关 键 词：岩体 可爆性 分级 投影寻踪 蚁群算法 

分 类 号：TD231.1[矿业工程—矿井建设] TD235.46