基于MIC-BBO-SVM的大坝渗流预测模型*

为监控大坝运行过程中的异常状态，准确预测大坝渗流量的变化趋势，采用最大信息系数(MIC)量化渗流量与影响因子之间的相关性大小并从中选取主导因子作为输入变量，通过引入生物地理学优化算法(BBO)并以K折交叉验证意义下的平均均方根误差为损失函数来优化支持向量机(SVM)作为预测模型，以某水电站工程的拦河大坝为例进行模型验证。结果表明:MIC-BBO-SVM模型的拟合优度、均方根误差、平均绝对误差和平均绝对百分比误差分别为0.957 5，0.155 0 m3/h，0.135 6 m3/h，11.51%，预测性能明显优于逐步回归模型、SVM模型和MIC-SVM模型，可为大坝渗流安全监测提供参考与借鉴。

Prediction model of dam seepage based on MIC-BBO-SVM

In order to monitor the abnormal state during the operation process of dam,and accurately predict the change trend of dam seepage flow,the maximal information coefficient (MIC) was used to quantify the correlation between seepage flow and influencing factors,and the dominant factor was selected from them as the input variable.The biogeography based optimization algorithm (BBO) was introduced and the average root mean square error in the sense of K fold cross validation was taken as the loss function to optimize the support vector machine (SVM) as a prediction model,then the dam of a hydropower station project was taken as an example for the model verification.The results showed that the goodness of fit,root mean square error,average absolute error and average absolute percentage error of the MIC-BBO-SVM model were 0.957 5,0.155 0 m3/h,0.135 6 m3/h,and 11.51%,respectively.Its prediction performance was better than those of the stepwise regression model,SVM model and MIC-SVM model,which can provide reference for the safety monitoring of dam seepage.