摘要
选择以对爆破效果影响最为明显的爆破参数作为神经网络输入层参数,建立神经网络预测模型,对爆破效果参数进行预测。将块度指标作为神经网络计算的输出层神经元,分析神经网络计算得出的块度指标是否满足设计精度。选择合理的样本数目进行训练学习,将模型计算结果与实测现场数据进行比较,输出结果的误差精度可以满足现场生产需要,能够达到指导实践中矿山调整爆破参数、改善爆破效果,节约成本的目的。
In order to predict the blasting parameters, the Neural network was used. The input neurons were set up by blasting parameters which were the most obvious influence on the blasting in the neural network prediction model. The block index was used as the output layer neuron of the neural network, analyzing of the neural network calculation of the block index to meet the design accuracy. The reasonable samples were selected and compared with the measured data on the spot. The results showed that the error precision of the output results could satisfy the requirement of the field production. Therefore, the prediction model of blasting parameters could be used to guide the adjustment of blasting parameters, to reduce the costs of productions.
出处
《辽宁科技大学学报》
CAS
2016年第4期303-306,310,共5页
Journal of University of Science and Technology Liaoning
基金
辽宁科技大学大学生创新创业训练计划(DC2015224)
关键词
爆破参数
神经网络
效果
预测
blasting
parameters
neural network
effect
forecast
