【摘要】 准确预测爆破振动速度对优化爆破参数和减少爆破产生的环境影响具有重要的意义。以某露天矿山生产爆破监测数据为例，采用通径分析理论确定了影响爆破振动速度的关键因素；结合Tent混沌映射优化SSA-BP神经网络初始种群位置的方法，建立了基于PA-Tent-SSA-BP的露天矿爆破振动速度预测模型。研究结果表明：与PSO-BP、GWO-BP以及SSA-BP神经网络预测模型相比，该模型的预测值和实测值更接近，RMSE、MAE以及MAPE分别为0.64、0.53以及0.18,说明该方法具有较好的泛化能力和预测性能，为多因素影响下爆破振动速度预测提供了一种新的研究思路。更多还原

【Abstract】 Accurately predicting blasting vibration velocity was of greatsignificance for optimizing blasting parameters and reducing environmental impact caused by blasting. Taking the production blasting monitoring data in an open-pit mine as an example, the key factors affecting blasting vibration velocity were determined with path analysis theory.The initial population position of SSA-BP neural network is optimized by Tent chaotic mapping.Finally, the prediction model of blasting vibration velocity in an open-pit mine based on PA-Tent-SSA-BP isestablished.The results show that compared with PSO-BP, GWO-BP and SSA-BP neural network prediction models, the predicted value and the measured value of PA-Tent-SSA-BP neural network are closer, with RMSE, MAE and MAPE being 0.64, 0.53 and 0.18 respectively.It shows that this method has good generalization ability and prediction performance, and provides a new research idea for the prediction of blasting vibration velocity under the influence of multiple factors.更多还原