GNSS网络实时动态模糊度解算病态性解决策略

Resolving strategies for ill-posed equation in ambiguity resolution of GNSS network real-time kinematics

针对GNSS网络实时动态(RTK)参考站间模糊度解算病态性问题,分析了病态性对模糊度浮点解影响,并基于无电离层组合解算模糊度基本模型,提出了改善模型病态性的两种策略。1参数选取策略:针对高仰角卫星,采用相对天顶对流层参数代替常规双天顶对流层参数设置,减少待估参数以改善病态性;2参数相关性优化策略:将GNSS卫星模糊度解算分为较易固定和较难固定两类,首先获取较易固定模糊度整数解,并反演天顶对流层延迟信息,再将该信息作为先验信息对较难固定模糊度解算模型进行约束,通过减小天顶对流层与模糊度相关性改善病态性。算例分析表明:两种策略在初始历元法方程病态性就明显优于常规模型,且只要通过少数十几个甚至几个历元就能够快速减弱法方程的病态性。该方法不需要考虑附加矩阵或参数的设置,易于实际工程应用。

The ill-posed problem in ambiguity resolution between reference stations in network RTK（real-time kinematic） has adverse influence on the float solution of the ambiguity. Based on the analysis on this influence and by using an ionosphere-free linear combination model, two strategies are put forward： One is the parameter-selecting strategy, which improves the ill-posed equation for satellites at high elevation by reducing the unknown parameters, that is, a single relative zenith tropospheric parameter was applied instead of conventional two zenith tropospheric parameters. The other is by optimizing the parameter＇s correlation, in which the ambiguities to be resolved are classified into two categories, one is easy to be fixed, and the other is difficult to be fixed. The former parameters are estimated and then inversely calculated to obtain the zenith tropospheric delay parameters, and then they are used as priori information to constrain the model. Thus the correlation between the zenith tropospheric parameters and the ambiguities is weakened, which improves the ill-posed equations. Test results for the two strategies show that the normal equations formed at initial epochs are evidently less ill-posed, compared with the conventional models. The ill-posed state of the normal equation can be rapidly weakened only by a few dozen or even several epochs. This method doesn＇t need to set the additional matrix or parameters, and can be easily applied into practical application.