将神经网络原理应用于河南灵宝罗山矿区中深孔爆破效果的预测中,选取爆区岩石力学性质、炸药性能参数、爆破设计参数作为影响爆破效果的因素,建立了神经网络模型,对爆破效果包括块度、大块率、前冲距离、堑沟边帮平整度、质点振动速度进行了非线性预测。结果表明,运用神经网络预测模型对爆破效果进行预测,取得了较好的效果,同时,可以根据预测的爆破效果调整爆破参数来满足工程要求。该方法对于节省工程投资和提高工作效率有一定的实用价值。
In the project of medium-depth hole blasting in Luoshan mining area, Henan Province, the theory of neural network was adopted in the nonlinear prediction of blasting effects including block size, block rate, forward setting distance, trench-boundary evenness, particle vibration velocity. In the analysis, rock mechanics properties, explosive performance and blasting design parameters were selected as the factors influencing blasting effects, based on which the neural network model was established. It shows a better result can be obtained in the prediction of blasting effects with neural network model. Furthermore, blasting parameters can be adjusted according to the prediction results so as to meet engineering requirements. This proposed evaluation method has certain practical value in saving costs and improving efficiency of projects.