Application of Improved BP Neural Network to GPS Height Conversion

The adjusted GPS height is the height above the WGS-84 ellipsoid. It is necessary to convert a GPS height into a normal height in engineering applications. GPS height conversion is usually used the standard BP (back-propagation algorithm) neural network model, but there are some defects in standard BP algorithm: low efficiency and easy to fall into local minimum. Aiming at overcoming the slow convergence rate and its encountering local minimum of traditional BP neural network, the paper establishes an improved BP neural network to GPS conversion, which based on LM algorithm and combination of momentum factor and adaptive learning rate. It is shown from the results of a practical engineering example that the improved BP neural network algorithm can significantly reduce the neural network training time and improve the efficiency of height transformation in GPS height.