提出采用选权拟合的正则化方法。利用参考站坐标准确已知的条件作为约束,设计出正则化矩阵,使法矩阵的病态性得到了较大的改善,只用几个历元的数据求解,能较准确得到模糊度的浮点解。在此基础上,结合LAMBDA方法可以快速地确定整周模糊度。以4条中长基线的实测数据为例,本文方法用5个历元的单频L1数据就能求解出整周模糊度。与现有方法相比,新方法可采用单频GPS数据实现中长基线GPS网络RTK的整数模糊度快速解算,而且可靠性好,有良好的应用前景。
A new method is proposed based on the Tikhonov regularization principle and selected weight fitting thought and precisely known reference station coordinates, to work out the regularized matrix, which can mitigate the ill-condition of the normal matrix. The relative float ambiguity can be computed only using several epochs' data. Combined with the LAMBDA method, the new approach can fix the integer ambiguity correctly and quickly. Taking an example, using measured data sets coming from the four medium-long baselines, the new method can obtain the exact ambiguities only using L1-frequency data of five epochs. Compared with the existing methods, the new method can solve the ambiguities of medium-long baseline GPS network RTK only using L1-frequency GPS data reliably. In practice, it can save the cost for fast positioning by GPS and would be applied widely.