基于组合优化灰色模型的尾矿坝浸润线预测

A Prediction of Saturation Line of Tailings Dam on a Combined Optimization Grey Model

为提高贫信息情况下浸润线的预测精度,减少尾矿坝预警的误报与漏报,提出了一种利用组合优化灰色预测模型对尾矿坝浸润线监测数据进行分析的方法。本方法对监测序列进行缓冲处理并基于非齐次指数函数对灰色预测模型的背景值进行了优化,采用后验比、小概率误差和平均相对模拟百分误差作为模型拟合预测精度评价指标,以实例计算验证所建模型的合理性,并与其他灰色预测模型的模拟预测结果进行对比,结果表明组合优化灰色预测模型拟合预测精度更高。模型为尾矿坝的安全监测与病险防护提供了数据支持,为大坝原型观测资料处理提供了新途径。

To improve the prediction accuracy of phreatic line under the condition of poor information and reduce the false alarm and missing alarm of tailings dam early warning,a method of analyzing the monitoring data of tailings dam phreatic line by using the combined optimization grey prediction model is proposed.In this method,the monitoring sequence is buffered and the background value of the grey prediction model is optimized on the non-homogeneous exponential function.The posterior ratio,small probability error and average relative simulation percentage error are used as the evaluation indexes of model fitting prediction accuracy.The rationality of the model is verified by example calculation in comparing with the simulation prediction results of other grey prediction models.The results show that the fitting prediction accuracy of this method is higher.This method provides data support for the safety monitoring and disease prevention of tailings dam,as provides a new way for the processing of dam prototype observation data.