基于退火BP神经网络的GPS高程转换

GPS height conversion based on Simulated Annealing BP Neural Network Model

阐述模拟退火算法的基本思想和原理,提出并介绍模拟退火算法优化的BP神经网络模型在GPS高程转换中的具体应用,同时编写相应的MATLAB处理程序,结合大量数据进行仿真实验,结果表明文中提出的退火BP神经网络模型具有收敛速度快、精度高、避免陷入局部最小的优良特性。

This article elaborates the basic thought and the principle of the simulated annealing algorithm, puts forward and introduces the application of BP Neural Network optimized by the Simulated Annealing Algorithm in GPS height conversion, compiles corresponding MATLAB programs, carries on a simulation experiment with mass GPS data. The result shows that the Simulated Annealing BP Neural Network has faster convergence rate, higher precision, and can avoid partial minimum effectively.