FOA优化GRNN网络的尾矿库安全预测

从尾矿库安全管理实际出发,针对尾矿库安全预测影响因素多、波动性大和非线性的特点,提出了果蝇算法优化广义回归神经网络的尾矿库安全预测模型。通过利用果蝇优化算法的全局寻优特性对广义回归神经网络进行参数优化,同时采用相关分析方法选取尾矿库安全评价指标,实现尾矿库的安全评价预测。以辽宁本溪南芬尾矿库为研究实例进行预测仿真,实验结果表明:相较于GRNN网络模型和BP网络模型,采用果蝇算法优化的GRNN模型预测精度更高,适用性更强,在尾矿库安全预测方面具有很大的实际应用价值。

Safety prediction of mine tailing reservoir based on FOA-GRNN algorithm

According to the actual situation of safety management on tailing reservoir , and aiming at the characteris-tics in safety prediction of tailing reservoir with many influencing factors , stochastic fluctuation and nonlinear , a safety prediction model for tailing reservoir was put forward by using generalized regression neural network optimized by fruit fly optimization algorithm .By using the global optimization characteristic of FOA for GRNN parameters opti-mization, the correlation analysis method was applied to select the safety evaluation indexes of mine tailings to a -chieve forecast .Taking Liaoning Benxi Nanfen tailing reservoir as research instance to conduct forecast and simula -tion, it showed that comparing with model of GRNN network and BP network , FOA-GRNN model has higher pre-diction precision and stronger applicability , and it has great practical application value in safety forecast of tailing reservoir .