大坝安全监测的遗传—偏回归(GA-PLSR)模型研究及应用

Study on Genetic Algorithm & Partial Least Square Regression Model for Dam's Safety Monitoring

基于在大坝监测中常用的偏回归模型的基础上引入了遗传算法,充分利用其强大的自适应全局优化概率型搜索功能,采用改进的遗传算法对偏回归系数进行寻优重估,建立大坝安全监测的遗传-偏回归模型。从而达到对偏回归模型优化的目的,以同时解决和改善常规大坝安全监测回归模型中存在的因子多重相关性干扰和模型欠拟合问题,进一步提高大坝监控模型的拟合和预测精度。

Here,the genetic algorithm is introduced based on the partial regression model commonly used in dam monitoring,at the same time, by using the powerful adaptive global optimization probabilistic searching function and using the improved genetic algorithm for optimization and re-evaluation of the partial regression coefficients,the genetic-partial regression model is established for dam safety monitoring.Then,the purpose of optimization for the partial regression model is achieved,and such problems as the multi-correlation interfereuce and model underfitting which exist in the general regression model for dam safety monitoring are also solved and improved,so as to further raise the fitting and prediction accuracy of the dam monitoring model.