基于 GA-BP 神经网络的爆破振动速度预测研究

打开文本图片集

DOI：10.19981/j.CN23-1581/G3.2026.11.009

中图分类号：TD235

文献标志码：A

文章编号：2095-2945（2026）11-0042-05

Abstract： Accurate prediction of blasting-induced peak particle velocity （PPV） is crucial to reducing the negative impact of blasting vibration. In order to improve the accuracy of PPV prediction， a convolutional neural network （CNN） prediction model， a BP neural network prediction model， and a genetic algorithm-based backpropagation optimization neural network （GA-BP） prediction model were constructed to predict vibration velocity. Based on the blasting mining monitoring data of Jianshan Phosphate Mine， relevant data were selected as input parameters， and the three prediction results were compared. The research results show that the prediction effect of the GA-BP neural network model is significantly better than that of the CNN neural network prediction model and the BP neural network prediction model. The R2 reaches 0.941， and the RMSE value is reduced by 31.25% and 16.41% respectively; the MAE value is reduced by 51.16% and 26.51% respectively， which verifies the superiority of the model. In the later stage， the model will be trained based on monitoring data from other mines to improve the generalization ability of the GA-BP neural network model， and will be implanted into a mine safety real-time monitoring and early warning platform through transfer learning.

Keywords： peak particle velocity （PPV）; vibration speed prediction; correlation analysis; BP neural network; genetic algorithm

爆破振动是矿山开采、隧道掘进等工程中不可避免的副产物，其引发的质点峰值振速过高可能导致周边建筑物损伤、岩体失稳及环境扰动 [1-2] ，严重威胁工程安全与居民生活。（剩余5412字）