摘要
建立合理的大坝变形预警模型对于大坝安全稳定运行意义重大。为提高预测精度,建立以相关向量机(RVM)为理论基础的时间序列非线性预警模型,采用一种精度较高的时间序列短期预测(自回归移动平均ARMA)模型修正RVM预测模型的误差序列,同时采用一种改进的粒子群算法(PSO)寻优核函数。实例验证结果表明,修正后的模型预测结果精度明显提高,可为类似工程提供参考。
The reasonable dam deformation early warning model is significance for safe and stable operation of the dam. In order to improve the prediction accuracy, nonlinear time series early warning model is established based on the relevance vector machine (RVM). A high accuracy time series short-term prediction method which called auto-regressive moving average (ARMA) model is used to correct the error series of the RVM prediction model. At the same time, the kernel parameter is optimized by an improved particle swarm optimization. Instance results demonstrate that the RVM- ARMA early warning model can improve the accuracy of deformation prediction effectively, which provides reference for the similar projects.
出处
《水电能源科学》
北大核心
2015年第7期89-91,38,共4页
Water Resources and Power
基金
国家自然科学基金重点项目(41323001
51139001)
高等学校博士学科点专项科研基金项目(20120094110005
20120094130003
20130094110010)
新世纪优秀人才支持计划资助项目(NCET-11-0628
NCET-10-0359)
水利部公益性行业科研专项经费项目(201201038
201301061)
作者简介
杜传阳(1990-),男,硕士研究生,研究方向为水工结构安全监控,E—mail:986952690@qq.com