摘要
为了有效地根据土石坝地原型观测资料来分析坝体和坝基中存在的渗流问题,本文在对建模因子进行分析的基础上,充分利用支持向量机的结构风险最小化原则和遗传算法快速全局优化的特点,通过支持向量机模型对非线性监测数据进行拟合,并利用遗传算法优化支持向量机的模型参数,建立了基于GA-SVM的土石坝渗流监测模型。实例分析表明,该模型与传统的多元线性回归模型和神经网络模型相比,具有预测精度高、泛化能力强等优点,对土石坝安全监控具有实用价值。
Based on the analysis on the factors influencing the safety of earth-rock dam the support vector machine is applied to testablish the model for simulating the seepage monitoring data of earth-rock dams, and the genetic optimization is used to optimize the parameters of the model. The application shows that this model is better than the models based on multiple element regression and neural networks, it possesses the advantage of high accuracy of forecasting and high ability of generalization. The calculation result shows that the seepage predicted by this model is in good agreement with the obsevation data.
出处
《水利学报》
EI
CSCD
北大核心
2007年第11期1341-1346,共6页
Journal of Hydraulic Engineering
关键词
土石坝
渗流安全监测
支持向量机
遗传算法
earth-rock dam
seepage safety monitoring
support vector machine
genetic algorithm
