基于RBF神经网络的露天矿爆破效果预测研究

Study on blasting effect prediction of open-pit mine based on RBF neural network

露天矿爆破是一个受诸多因素共同影响的系统工程,是露天开采的重要环节之一,其爆破效果的优劣直接影响后续工序的完成。提高爆破技术水平和爆破质量,对矿山安全和生产具有重要的意义。本文通过随机森林选择影响爆破效果的主要参数,结合模糊评价确定爆破综合效果,建立了RBF神经网络爆破效果预测模型。将该模型应用于矿山爆破效果预测中,并将爆破现场实测的11组数据作为模型训练样本,另外5组现场数据作为预测样本进行测试,通过与BP神经网络比较,发现RBF神经网络的预测性能更为优越,可广泛应用于现场实践中。

Open-pit mine blasting is a system engineering affected by many factors,and it is one of the important links of open-pit mining.Its blasting effect directly affects the completion of the follow-up process.Improving blasting technology and quality is of great significance to mine safety and production.In this paper,the main parameters affecting blasting effect are selected by random forest,and the comprehensive blasting effect is determined by fuzzy evaluation.The prediction model of blasting effect based on RBF neural network is established.The model is applied to mine blasting effect prediction.Eleven groups of data measured in the blasting site are used as training samples,and five groups of data are used as prediction samples to test.Compared with BP neural network,it is found that RBF neural network has better prediction performance and can be widely used in field practice.