论文部分内容阅读
结合粗糙集理论与遗传算法,提出一种新的岩爆倾向性预测方法。以碳化矿床采场工作面岩爆倾向性预测为例,选取9种岩爆影响因素作为条件属性,将采场工作面岩爆倾向性结果作为决策属性,选取77组采场岩爆或非岩爆工程实例数据,构成岩爆倾向性预测样本集合。将样本集合随机分为训练样本和测试样本2个集合,对2个集合的属性值进行量化,建立训练样本决策表与测试样本决策表。采用遗传算法对训练样本决策表的条件属性进行约简,得出条件属性的最小约简和核。应用粗糙集理论从约简结果中提取精简的岩爆倾向性判别规则集。用获得的规则集对测试样本决策表中的岩爆倾向性进行预测,并与实际结果对比,验证了判别规则集的可行性和有效性。结果表明,基于粗糙集与遗传算法的岩爆倾向性预测方法更具客观性和科学性,有较高的工程实用价值。
Combined with rough set theory and genetic algorithm, a new method of rockburst prediction is proposed. Taking the tendency prediction of rock burst in stope face of carbonitized ore deposit as an example, the influence factors of nine kinds of rockburst are selected as condition attributes. Taking the rock burst tendency of stope as a decision attribute, 77 rockburst or non- Explosion project instance data, constitute a set of predictions of rockburst samples. The sample set is randomly divided into two sets of training samples and test samples, and the attribute values of two sets are quantified to establish the decision table of the training samples and the decision table of the test samples. The genetic algorithm is used to reduce the condition attributes of decision table of training samples, and the minimum reduction and kernel of condition attributes are obtained. Apply Rough Set Theory to Extract Reduced Rockburst Probability Discrimination Rule Set from Reduction Results. The obtained rule sets are used to predict the rockburst tendencies in the test sample decision table. Compared with the actual results, the feasibility and validity of the discriminant rule sets are verified. The results show that the method of rockburst prediction based on rough sets and genetic algorithm is more objective and scientific, and has higher engineering practical value.