基于EEMD-LSTM模型的尾矿库风险预测模型研究*

为了提高尾矿库风险预警能力，针对尾矿库稳定性受多种风险因素影响，以及风险变化的非线性，提出1种融合集合经验模态分解（EEMD）和长短期记忆（LSTM）的尾矿库风险预测模型。首先，采用皮尔逊相关系数分析尾矿库风险因素之间的相关性；然后，使用EEMD方法分解非线性的位移序列；最后，构建LSTM网络模型预测位移变化。以某尾矿库为例，将EEMD-LSTM模型与EEMD-BP模型、LSTM模型对比分析，评估模型的有效性。研究结果表明:EEMD-LSTM模型对尾矿库风险的预测精度明显提高，对防范化解尾矿库安全风险具有重要意义。

Research on risk prediction model of tailings pond based on EEMD-LSTM model

In order to improve the risk early-warning capability of tailings ponds,and considering that the stability of tailings pond is affected by multiple risk factors,as well as the nonlinearity of risk change,a risk prediction model of tailings pond with the fusion of ensemble empirical mode decomposition (EEMD) and long short-term memory (LSTM) was proposed.Firstly,the Pearson correlation coefficient was used to analyze the correlation between the risk factors of tailings pond.Then,the EEMD method was used to decompose the nonlinear displacement sequence.Finally,the LSTM network model was constructed to predict the displacement change.Taking a tailings pond as an example,the EEMD-LSTM model was compared with the EEMD-BP model and the LSTM model to evaluate the effectiveness of the model.The results showed that the EEMD-LSTM model could significantly improve the prediction accuracy of tailings pond risk,which is of great significance for preventing and resolving the safety risk of tailings pond.