论文部分内容阅读
为预防和减少露天铁矿爆破过程中产生的超爆和飞石引起的灾害,必须改进现有爆破方案。考虑到爆破方案参数包括:炮眼深A、间距B、装药深度C、阻塞深度D、单位炸药消耗量E和钻孔率F。它们对超爆深度G和飞石距离H的影响是复杂且模糊的。提出用结构元直觉模糊集改进的遗传算法GA来选择最优爆破方案。用结构元方法表述模糊数直觉模糊集,划分出好、犹豫、不好等模糊语,对爆破方案参数进行模糊化。按照实际统计特征,随机生成100个爆破方案。以用TOPSIS方法构造的函数,作为GA的适函数。根据超爆深度和飞石距离同时达到最优的准则,对100个方案进行排序。考虑与实际的关联问题,使用神经网络筛选子代中G和H,保留符合要求的方案,并进行淘汰方案的数量补充。试验结果表明,结构元直觉模糊集改进的GA算法提高了收敛性和适应性,得到的最优爆破方案符合实际情况。
In order to prevent and reduce the disasters caused by the over-blasting and fly-rock during the blasting of the open-air iron mine, the existing blasting program must be improved. Taking into account the blasting program parameters include: blasthole depth A, spacing B, charge depth C, blocking depth D, unit explosive consumption E and drilling rate F. Their effect on the depth of hypersonic G and the distance H from the fly is complex and ambiguous. The improved genetic algorithm GA based on structural element intuitionistic fuzzy set is proposed to select the optimal blasting scheme. Structural element method is used to express the intuitionistic fuzzy set of fuzzy numbers, and fuzzy words such as good, hesitant and bad are divided to blur the parameters of the blasting scheme. In accordance with the actual statistical characteristics, a random generation of 100 blasting program. Functions constructed using the TOPSIS method serve as the fitness function for GA. According to the super-blast depth and fly-distance at the same time reach the optimal criteria, the 100 programs are sorted. Consider the actual association problem, use neural network to screen G and H in the offspring, keep the scheme that meets the requirement, and add the number of phase-out scheme. The experimental results show that the improved GA algorithm of structural element intuitionistic fuzzy set improves the convergence and adaptability, and the optimal blasting scheme accords with the actual situation.