Gps Height Measurement Conversion Method

In the foundation of the elaboration elevation system and the GPS height measurement principle, this article analyzes the affecting factor of GPS height measurement and the GPS geodetic height measurement precision. Then makes a comprehensive and careful analysis to the existing GPS height measurement method , applying to a project example , studies the using range and the precision of different fitting models . Aiming at the insufficiency of the geometry method, the author studies a new GPS height conversion method which bases on the artificial neuron network, and proves the feasibility of the artificial neural networks conversion the GPS height through analyzing the example, also discusses determination of the implicit strata unit quantity, the concealment layer determination, the study speed choice, the initial weight choice, and the training sample exudes ability questions and so on influence in the BP network architecture . To avoid the limitation of applying the sole model to carry on the elevation fitting method, relying on absorbing and studying the research results, the author makes different fitting model spliced, so as to improve the height anomaly approaching precision and the reliability.