摘要
大坝渗流监测分析是大坝安全监控的重要内容,预测分析的难点之一在于渗流监测数据往往具有复杂的非线性特点。本文充分利用支持向量机的结构风险最小化与粒子群算法快速全局优化的特点,采用粒子群算法快速优化支持向量机的模型参数,通过该模型对非线性监测数据进行拟合,建立了基于PSO_SVM的大坝渗流监测的时间序列非线性预报模型。本模型应用于隔河岩水电站的坝基渗流量的预测,计算结果与实际监测值吻合良好。
The support vector machine based on statistical learning theory is applied to establish a time series model for simulating the monitoring data of dam seepage flow. The particle swarm optimization is used to optimize the parameters of the model. On this basis, the nonlinear time series prediction model for dam seepage flow based on PSO-SVM program is established. The principle and working steps of the method are presented. The seepage predicted by this method based on the monitoring data for the dam in Geheyan Project is in good agreement with the observation data.
出处
《水利学报》
EI
CSCD
北大核心
2006年第3期331-335,共5页
Journal of Hydraulic Engineering
关键词
安全监测
粒子群算法
支持向量机
时间序列预测
dam safety monitoring
particle swarm optimization
support vector machine
time-series prediction
