摘要
介绍了利用双通道地基微波辐射计,基于神经网络和遗传算法结合的晴空大气剖面反演研究工作,神经网络算法反演折射率干项,遗传算法反演折射率湿项,使用青岛地区历史探空数据仿真表明,两种算法相结合反演的大气折射率剖面与探空数据吻合较好。
A new method of atmospheric refractivity profile retrieval based on neural network and genetic algorithm by groundbased dual-channel radiometer was proposed. Retrieval of dry refractivity profile was based on neural network and wet refractivity profile based on genetic algorithm. Simulations with the historical radiosonde data of Qingdao showed that the new retrieval method has good agreement with historical atmospheric profiles.
出处
《装备环境工程》
CAS
2008年第1期53-55,共3页
Equipment Environmental Engineering
基金
国家973项目
电波环境特性及模化技术国家重点实验室基金(9140C0802050703)资助
关键词
微波辐射计
神经网络
遗传算法
大气折射率
radiometer
neural network
genetic algorithm
atmospheric refractivity
