为提高传统大坝位移统计模型的预测能力,引入蛙跳算法和卡尔曼滤波等,基于最小二乘法确定统计模型回归系数,采用蛙跳算法对回归系数进行优化;运用卡尔曼滤波法将回归系数作为状态方程的状态向量,使预测模型能实时反映监测数据的变化。工程算例应用结果表明,基于蛙跳算法与卡尔曼滤波的大坝位移预测优化模型能有效提高统计模型的预测精度,可为大坝位移预测提供参考。 In order to improve the predictive ability of the traditional dam displacement statistical model, the frog leaping algorithm and Kalman filter are introduced to determine the regression coefficient of the statistical model based on the least square method, and the frog jump algorithm is used to optimize the regression coefficient. The Kalman filter is used to divide the regression coefficient As the state vector of the state equation, the prediction model can reflect the change of the monitoring data in real time. The results of engineering application show that the dam displacement prediction optimization model based on leapfrogging algorithm and Kalman filter can effectively improve the prediction accuracy of the statistical model and provide a reference for the dam displacement prediction.