基于集成学习CatBoost优化模型的爆堆大块率预测

Prediction of Blasting Fragment Large Block Percentage Ratio Based on Ensemble Learning CatBoost Model

爆破产生的爆堆大块率问题一直以来都影响着矿山的生产效益.利用机器学习机制中集成学习思想实现大块率预测.以满洲里乌山铜矿实际采集的36组实测数据为例,整理形成10种特征数据.通过给定参数循环训练调优,再用交叉验证网格搜索的方法进行模型二次调优,并对比调优实现后模型与随机森林法、XGBoost模型、LightGBM模型和CatBoost模型进行效果对比.结果表明,经过两轮调优后的CatBoost模型预测效果明显高于其他几种模型,R~2准确度可达98.83%,证明了两轮调优后CatBoost模型具有较高的预测水平,验证了该方法在大块率预测研究中的可行性,为爆破参数设计和大块率优化分析提供了可靠的参考.

The problem of large block percentage ratio of blasting has always affected the production benefit of mine.The implementation of large block percentage ratio of blasting predication based on ensemble learning in machine learning is studied.Firstly,36 groups of measured data collected from Wushan Copper Mine in Manzhouli are utilized as an example,and 10 kinds of characteristic data are sorted out.Then,the model is trained and tuned by cyclic training of the given parameters,and then the method of cross-validation grid search is used to perform secondary tuning,and the effect of the model after tuning is compared with random forest method,XGBoost model,LightGBM model,and CatBoost model.The results show that the prediction effect of CatBoost model,after two rounds of tuning,is significantly higher than that of other models.The precision rate 98.83% reaches,which proves that CatBoost model has a high prediction level after two rounds of tuning.At the same time,the feasibility of this method in the study of large block percentage ratio of blasting prediction is proved,which provides a reliable reference for blasting parameter design and large block percentage ratio of blasting optimization analysis.