Combining the genetic algorithms with BP Neural Network for GPS height Conversion

Plan control results of GPS surveying have been widely used in all kinds of engineering, while its height information is being studied at present. Because GPS height is the height above the WGS-84 ellipsoid, however, the normal height, which is the height above the geoid calculated by using the mean normal gravity along the plumb line, is used in engineering applications. It is necessary to convert GPS height into normal height. GPS height conversion is usually used the BP (back-propagation) neural network model, but there are some defects in BP algorithm. Aiming at overcoming the slow convergence rate and its encountering local minimum of traditional BP neural network, this paper introduces the GA (genetic algorithms), and proposes a new method, combining the genetic algorithms with BP Neural Network for GPS height Conversion. Based on real GPS surveying datum, we did an experiment with this method to GPS height. The compared and analyzed test results show that the combining the GA with BP neural network for GPS height conversion can achieved higher precision. At the same time, this method can significantly settle many questions that BP neural network must face.