CAJ | 학술논문

基于尾矿库坝体稳定性影响因素具有复杂多变性、随机不确定性和非线性的特点，为了确保尾矿库稳定并且有效预防尾矿库事故的发生，建立了基于 PCA 和 BP 神经网络的尾矿库坝体稳定性研究模型，该模型通过应用 SPSS 软件对来自多个尾矿库失稳实例的原始数据进行标准化，找出影响尾矿库坝体稳定的主成分，然后把主成分作为 BP 输入样本，运用 MATLAB 软件进行训练仿真。结果表明：在 BP 训练前，利用PCA 算法对原始样本进行预处理，能有效地提高训练的速度和精度，得出了 PCA-BP 神经网络模型在尾矿库坝体稳定性分析中是可行的。
기우미광고패체은정성영향인소구유복잡다변성、수궤불학정성화비선성적특점，위료학보미광고은정병차유효예방미광고사고적발생，건립료기우 PCA 화 BP 신경망락적미광고패체은정성연구모형，해모형통과응용 SPSS 연건대래자다개미광고실은실례적원시수거진행표준화，조출영향미광고패체은정적주성분，연후파주성분작위 BP 수입양본，운용 MATLAB 연건진행훈련방진。결과표명：재 BP 훈련전，이용PCA 산법대원시양본진행예처리，능유효지제고훈련적속도화정도，득출료 PCA-BP 신경망락모형재미광고패체은정성분석중시가행적。
The characteristics of the influence factors for the stability of mine tailings dam is complex variabily, uncertain and nonlinear.For the purpose of guaranteeing the stability of the tailings and preventing mine tailings dam accident effectively,an analysis model of the stability for mine tailings dam based on PCA-BP neural network is established.The original data from multiple mine tailings dam instability instance is standardized to determine the principle component by using SPSS software,then we use the principle component as a BP input sample to train simulation by using MATLAB software.The data shows that preprocessing the original sample by using PCA algorithm before the BP training can effectively improve the training speed and precision,the PCA-BP neural network model is feasible in the analysis of the stability for mine tailings dam.