基于SMPSO-SVM的土石坝运行期渗流监测模型的建立与运用

Establishment and application of seepage monitoring model for earth-rock dam during operation based on SMPSO-SVM

将SM算法嵌入PSO算法当中,在支持向量机模型上,建立关于土石坝安全运行的渗流监测模型(SMPSO-SVM),避免了SVM参数的随意选择性。与其它监测模型相比, SMPSO-SVM引入影响因子更少,大大降低了计算的复杂程度,且拟合预测的精度较高,计算过程更加稳定;模型运用到实际土石坝工程的渗流监测中,取得了不错的分析效果,可为类似土石坝的渗流安全监测提供新的方法和手段。

SM algorithm is embedded in PSO algorithm,and a seepage monitoring model(SMPSO-SVM)for the safe operation of earth-rock dams is established on the support vector machine model,which avoids the random selection of SVM parameters.Compared with other monitoring models,SMPSO-SVM has fewer influencing factors,greatly reduces the complexity of calculation,and has higher accuracy of fitting prediction and more stable calculation process.The model has been applied to seepage monitoring of earth-rock dam projects,and achieved good analysis results,which can provide new methods and means for seepage safety monitoring of similar earth-rock dams.