摘要
为准确预测矿井粉尘浓度,有效防治矿井粉尘危害,运用遗传算法优化的BP神经网络预测模型(GA-BP模型)对某矿山工作面时间序列粉尘浓度进行预测,以预测结果的相对误差、平均绝对百分比误差来评判模型的预测准确性。再利用BP神经网络预测模型、卷积神经网络预测模型(CNN模型)的预测结果同GA-BP预测模型的预测结果进行对比验证,以均方根误差来评价三种模型的预测效果。结果表明,应用GA-BP预测模型,相对误差最大为4.27%,最小为0.14%,相对误差都在10%以内,预测样本的平均绝对百分比误差(MAPE)小于10%,达到了高精度预测要求。CNN、BP、GA-BP三种预测模型的RMSE值分别为1.1007、1.0008、0.9354,GA-BP预测模型对于该矿山工作面粉尘浓度预测效果最好。实现了利用GA-BP神经网络预测模型对只有单一时间影响因素且样本数量较少条件下的矿井粉尘浓度预测。
In order to accurately predict the mine dust concentration and effectively prevent and control the mine dust hazard,the BP neural network prediction model(GA-BP model)optimized by the genetic algorithm was used to predict the dust concentration in a mine working face time series.The prediction accuracy of the model was evaluated by the relative error and the average absolute percentage error of the prediction results.By using the BP neural network prediction model,the prediction results of the convolutional neural network prediction model(CNN model)and the GA-BP prediction model were compared and verified,and the root-mean-square error was used to evaluate the prediction effect of the three models.The results show that the maximum relative error of the GA-BP prediction model is 4.27%,the minimum is 0.14%,the relative error is less than 10%,the average absolute percentage error(MAPE)of the predicted sample is less than 10%,which meets the requirement of high precision prediction.The RMSE values of CNN,BP,and GA-BP are 1.1007,1.0008,and 0.9354,respectively.The GA-BP prediction model has the best effect on the dust concentration prediction of the mine working face.The use of the GA-BP neural network prediction model to predict mine dust concentration under the condition of only a single time influencing factor and a small sample size is realized.
作者
周昌微
谢贤平
都喜东
ZHOU Changwei;XIE Xianping;DU Xidong(Faculty of Land Resources Engineering,Kunming University of Science and Technology,Kunming 650093,China)
出处
《有色金属（矿山部分）》
2023年第6期88-93,共6页
NONFERROUS METALS(Mining Section）
基金
云南省基础研究计划项目(202101BE070001-039)
云南省教育厅科学研究基金项目(2022J0055)。
