基于遗传神经网络的大气折光系数实时改正

Because of the slow convergence rate ,local minimum dot and some other questions,the neural network which combined with the union genetic algorithm has the overall situation optimization characteristic partially,and adopted in the total station triangle elevation atmospheric diffraction coefficient real-time correction, established the genetic neural network atmosphere diffraction coefficient correction model, which is more scientific and easy to realize. The real-time forecast of the atmospheric diffraction coefficient is carried on by this model according to the follwing four targets : terrain, the time interval, the temperature and the barometric pressure . Implementing the effective atmospheric diffraction coefficient real-time correction becomes the key to improve the precise of Total Station triangle elevation.