用遗传算法搜索GPS单频单历元整周模糊度

Searching Integer Ambiguities in Single Frequency Single Epoch by Genetic Algorithm

介绍了短基线利用单频单历元双差载波相位定位时模糊度固定的基本理论,探讨了利用遗传算法快速搜索GPS单频单历元整周模糊度的一些理论和实现的方法,提出了用改进的正则化方法改善浮动解来提高搜索成功率的新思路。算例分析表明,在一定的条件下,应用遗传算法搜索整周模糊度成功率高、稳键性较好。

Carrier phase ambiguity solution for single frequency single epoch is the key to high precision for real-time kinematic positioning. An evolution algorithm - Genetic Algorithm was introduced to search GPS carrier phase integer ambiguities in single frequency single epoch. It is a new optimization method through simulating the process of nature evolution to search the best solution. An improved regularization method is applied to improve the accuracy of the float solution of the ambiguities. The improved float solution increases the success rate of fixing the carrier phase integer ambiguities, since not only the candidate integer set is reduced, but also the ambiguity float solution is improved. An example is illustrated to show the new method's effect. The result processed by Bernese GPS software using all GPS data in this example is taken as the standard of comparison. Then, some L1 phase observations and C/A pseudo ranges are extracted, and each epoch's float solution is computed by the new regularization method. After obtaining more accurate ambiguity float solution, we could think ambiguity true value is around float solution for 3 circles. Under this precondition, Genetic Algorithm was used to search carrier phase integer ambiguity. The results indicated that the new method not only increased the success rate of fixing ambiguity, but also improved the search efficiency. Genetic Algorithm spent less time in searching integer ambiguity than that LAMBDA did. And Genetic algorithm was also proved a robust algorithm by some analyses using GPS data. The new research way based on the improving float solution will make great application significance to real-time kinematic positioning.