Neural network integrated navigation method

The invention discloses a neural network integrated navigation method. The method comprises the following steps of enabling signals to correspond, decomposing the signals of a navigation system, processing decomposition coefficients, forecasting and estimating a neural network of errors, building information of the errors of the navigation system, and correcting integrated navigation information. According to the neural network integrated navigation method, observed quantity which objectively exists is completely processed, a model is not related to, calculated quantity is small, and limiting of noise does not exist. Basically, an objective state is corrected by means of objective observation, correcting ability is strong, and an abstract state is not affected and controlled.

[1]  Xiyuan Chen,et al.  A new method based on WMRA and ANN for GPS/SINS integration for aerocraft navigation , 2006, International Symposium on Precision Mechanical Measurements.