摘要
遗传算法(GA)处理数值优化计算问题具有的简单通用、并行、稳健等特点,因此应用于高精度GPS定位的基线解算过程。针对双差模糊度的整数域和基线向量的实数域解的特性,进行了GA算法改进,包括实数编码的改进、遗传算子及其控制参数等算法设计,提出了基于非线性最小二乘准则的GPS相对定位同步解算基线向量和双差模糊度的优化搜索新方法,避免了分步解算模糊度中对浮点解的依赖性,首次实现了大范围、高精度、整数实数不同域上的同步求解,提高了GPS相对定位的稳定性,也体现了遗传算法的优越性。算例表明改进的实数编码遗传算法对同步解算GPS相对定位是可行有效的。
Because of some advantages, such as simpleness, parallel and robustness on resolving numerical value optimization problems, genetic algorithms （CA） were improved and applied on GPS baseline resolution. Aimed at the integer nature of double difference ambiguities and the real nature of baseline coordinates, the coded methods of CA were improved in order to satisfy the solu- tion sets characteristic of GPS positioning. And then the corresponding genetic operators and control parameters were modified. A method to solve the GPS relative positioning synchronously was proposed based on non - Linear least - square principle. So the dependence on the accuracy of float solution was avoided, and the improved CA helped to enhance the search algorithm success rate. Through a large number of eases, the tests of the proposed method were practiced and it was verified that this improved GA were superior to CPS carrier phase relative positioning resolution on stability and efficiency.
出处
《测绘科学》
CSCD
北大核心
2008年第5期135-136,217,共3页
Science of Surveying and Mapping
基金
国家自然科学基金(40704001)项目
关键词
改进的遗传算法
GPS载波相位相对定位
非线性最小二乘
改进的实数编码
the improved genetic algorithms
GPS relative positioning on carrier phase
non - linear least - square principle
the improved - real code
