Research On The Forecast Of Space Moving Object

The forecast tracking of the space moving object is a hot way of research in the field of military guidance. It is also the key technology of precision strike guided weapons. As to the forecast tracking, the trajectory of the moving object is very complex. It is very difficult to get the analytical expression of precise position. So the model would not accord with the real motion properties. And it is hard to choose forecast model and method too. The disaccord of forecast model and object real trajectory makes the forecast problem difficult.This paper bases on the path synthesis position forecast algorithm theory, changes the fast object racking problem to search optimized parameters of forecast model, and gives the fast object tracking method. It gives the object forecast strategy description that the forecast divides into two kinds, the far distance object forecast and near distance object forecast based on the BP neural network and space moving object tracking theory. The method dividing the object state by degrees of freedom and forecasting by subnet is proposed. Analyzed the feature of moving object, an improved immune algorithm with antigen evolution is proposed. The current information is used as the initial antigen and next one is as the antigen evolution in the optimization searching algorithm. An improved synthesis position forecast model is induced. Judging the antigen whether evolved and searching the best model parameters make the model get closer to the real one. Moreover, uses the optimized results to synthesize the position. Correctness and validity of the neural network object forecast algorithm and immune algorithm with evaluated antigen are tested by the simulation of varies examples.