摘要
从体系结构、硬件构成、网络结构和软件设计等几个方面来看,大坝监测自动化系统采集的数据会失真,针对以往监测数据验证方法存在的缺陷,提出了基于相空间重构和贝叶斯框架最小二乘支持向量机(BLS-SVM)相结合的监测物理量数据验证方法。采用LS-SVM构建了3种预测器,应用于单监测物理量和多监测物理量输出系统。实例结果证明了所提出的方案的有效性。
From the architecture, hardware, network architecture and software design, the data acquired is to be distorted. In view of the liraitation of previous data validation method, a new monitoring data validation method is proposed based on phase space restructuring and Bayesian least square support vector machines （BLS-SVM）. LS-SVM is used to construct three kinds of predictors which are applied to single variable and multi-variable output system. Experimental results show that the validation method is valid.
出处
《水电自动化与大坝监测》
2009年第3期46-50,共5页
HYDROPOWER AUTOMATION AND DAM MONITORING
关键词
大坝安全监测
数据验证
贝叶斯框架
最小二乘支持向量机
dam safety monitoring
data validation
Bayesian framework
least squares support vector machines （LS-SVM）
