CAJ | 학술논문

为改善地面重力数据缺乏区域似大地水准面的精度和分辨率，提出利用神经网络融合卫星重力信息应用于似大地水准面精化的方法，通过融合不同时期CHAMP卫星数据计算的重力场模型和GPS/水准数据中区域重力场信息，进行区域似大地水准面精化并进行精度统计，结果表明：融合卫星重力信息将区域似大地水准面模型的精度由0.4702 m提高至0.2038 m，验证了神经网络技术融合卫星重力信息并应用于区域似大地水准面精化是有效的．
위개선지면중력수거결핍구역사대지수준면적정도화분변솔，제출이용신경망락융합위성중력신식응용우사대지수준면정화적방법，통과융합불동시기CHAMP위성수거계산적중력장모형화GPS/수준수거중구역중력장신식，진행구역사대지수준면정화병진행정도통계，결과표명：융합위성중력신식장구역사대지수준면모형적정도유0.4702 m제고지0.2038 m，험증료신경망락기술융합위성중력신식병응용우구역사대지수준면정화시유효적．
In order to improve the precision and resolution of quasi-geoid in the region of lacking ground gravity data, the method of satellite gravity information fusion using neural network to refine regional quasi-geoid is proposed.The regional quasi-geoid is refined by fusing the gravity informa-tion of CHAMP ( challenging mini-satellite payload ) gravity field models at different periods and GPS/leveling, and then the precision of quasi-geoid is estimated.The results show that the satellite gravity information fusion can improve the precision of regional quasi-geoid from 0.470 2 m to 0.203 8 m, which verify that satellite gravity information fusion using neural network to refine re-gional quasi-geoid is effective.