摘要
AR模型在变形监测及其预报中应用广泛,因为系数矩阵和预测值都存在误差,宜采用整体最小二乘法。本文通过一个实例对该方法进行验证,计算结果表明,该方法获得的预测值更加接近实测值,并且比传统的方法更具优势,有较强的理论研究价值和实用价值。
AR model is widely used in deformation monitoring and forecasting, because the coefficient matrix and predictive value both exist errors, the total least-squares should be adopted. In the paper, it used an example to verify this method, and the results showed that the predictive value was much closer to the measured values, and had more advantages than the traditional method. It would have a strong theoretical study and practical value.
出处
《测绘科学》
CSCD
北大核心
2013年第2期171-172,167,共3页
Science of Surveying and Mapping
关键词
AR模型
变形监测
整体最小二乘
参数
AR model
deformation monitoring
total least-squares
parameter
