摘要
鉴于欧拉矢量参数在局部区域适应性较差,提出一种基于Elman神经网络的速度场逼近方法。首先利用已有欧拉矢量参数估计站点速度,将剩余残差作为Elman神经网络拟合量进行逼近;然后将Elman神经网络估计结果与欧拉矢量计算速度相叠加,获得区域速度场模型。利用山东区域速度场数据进行验证,结果表明,该方法在一定程度上能够削弱系统误差影响,提高计算精度。
The parameter of Euler vector has poor validity in a local area. An algorithm based on El- man neural network is introduced to fit the velocity field. First, the velocity of the position is compu- ted with the parameter of Euler vector; second, the residual as the expectation of Elman neural net- work is again trained; finally, we determine the velocity of the position, which is equal to the sum of the results obtained by Elman neural network and the velocity of Euler vector. The data set of Shan- dong is employed to test the algorithm. It is shown that the new algorithm can weaken the influence of systemic error and improve the accuracy of velocity field.
作者
聂建亮
郭春喜
曾安敏
田婕
张海平
王斌
程传录
NIE Jianliang GUO Chunxi ZENG Anmin TIAN Jie ZHANG Haiping WANG Bin CHENG Chuanlu(Center of Geodetic Data Processing, NASMG, 334 East-Youyi Road, Xi＇an 710054, China Xi＇an Research Institute of Surveying and Mapping, 1 Mid-Yanta Road, Xi＇an 710054, China Institute of Land Surveying and Mapping of Shandong Province, 2301 Jingshi Road, Ji＇nan 250102, China)
出处
《大地测量与地球动力学》
CSCD
北大核心
2017年第10期1015-1019,共5页
Journal of Geodesy and Geodynamics
基金
国家自然科学基金(41574003
41774004
41474015
41604001)
测绘地理信息公益性行业科研专项(201512004)
国家科技部科技基础性工作专项(2015FY210400)
现代工程测量国家测绘地理信息局重点实验室开放基金(TJES1504)
地理空间信息工程国家测绘地理信息局重点实验室开放基金~~
