摘要
由于影响大坝安全的客观因素非常复杂 ,大坝安全监控回归模型在实际工程应用中普遍存在欠拟合问题。为解决该问题 ,在回归模型的基础上引入改进的遗传算法 ,利用其自适应全局优化搜索功能对回归模型中的回归系数进行寻优重估 ,并建立遗传回归模型。以东北地区某宽缝重力坝为例 ,分别建立了其坝顶水平位移安全监控混合模型和遗传回归模型 ,计算结果表明 。
On the basis of regression model,the genetic algorithm is applied to improve the establishment of mathemiatical model for dam safety monitoring.It helps to overcome the under fitting problem usually occurred in regression.The application of this method to a slotted gravity dam located in the northeast of China is given as an example for demonstration.Hybrid model and genetic regression model for displacement of the top of the dam are established respectivelty.The calculation shows that by using this new method the fitting is greatly promoted.
出处
《水利学报》
EI
CSCD
北大核心
2002年第8期91-95,共5页
Journal of Hydraulic Engineering
关键词
大坝
监测数学模型
欠拟合
遗传回归模型
dam safety monitoring
mathematical model
under fitting
genetic regression model
