Application of PSO-BP Neural Network in GPS Height Fitting

Abstract

In order to improve the accuracy of GPS height fitting, aiming at the shortcomings of BP neural network in application, such as slow learning convergence speed and easy to fall into local optimal solution, particle swarm optimization (PSO) has the advantages of global search performance and fast convergence speed. Globally optimizing the initial connection weight and threshold of BP neural network, PSO Optimized BP neural network is served as a foundation of a GPS elevation fitting model. It is the results that demonstrate that the BP neural network optimized by PSO proves feasible for GPS height fitting as well as the fitting accuracy can be effectively improved, with the supply of definite reference value for the establishment of a high-precision GPS elevation fitting model.