Study On Parameter Identification And Impact Points Prediction For Trajectory Correction Projectiles

The trajectory correction projectile is a kind of projectiles that between ordinary projectile and missile. It is a classical integrate of ordinary projectile and modern high technology. It timely measure the trajectory parameters of specific trajectory, then deals with the measurement data, lastly generate control command that control projectile towards the intended target. Among this process, data handling is the core content of trajectory correction. In this paper, trajectory parameters identification and impact point prediction technique belong to the process of data handling is researched.With the view that aerodynamic parameters are the main affection of trajectory, aerodynamic parameters can be identified and estimated instantly in this paper. Synthesizes the mount and kind of measured ballistic data, instantaneity and accuracy of identified result, establish equation of state based on extend particle trajectory model and linear measurement equation based on GPS data. Choose extended kalman filter(memory attenuated mixed kalman filter and maximum likelihood adaptive kalman filter)、Sigma-point kalman filter(unscented kalman filter and central difference kalman filter) and cubature kalman filter to identify demand parameters, take a numerical simulation of each means, then compare and analyze the simulation results. Simulations reveal that:memory attenuated mixed kalman filter can greatly decrease on-line calculation in a proper accuracy range, improve the instantaneity of identification process; maximum likelihood adaptive kalman filter can reduce model error, improve estimation accuracy; the adopt of single row sampling can decrease the calculation of unscented kalman filter, improve the real-time of algorithm; the central difference kalman filter use the way of difference to solve nonlinear transfer, therefore it needn’t calculate the Jacobian matrix with high precision, but the calculation is large owing to the large quantity of sampling and its instantaneity is poor; cubature kalman filter adopt the ideas of cubature which use 2n equally weighted cubature point to solve nonlinear transfer, it possesses the advantages of easy to achieve, high accuracy and stability.After the identification process, it is necessity to predict the trajectory falling points to provide the basis for calculating trajectory control. In this paper, calculation prediction method based on trajectory formula and correction factor algorithm method are discussed. The comparison and analysis of the two prediction algorithm and numerical simulation are showed. The result showed that:the former has high precision and good adaptability, but its parameter requirements are very strict and its instantaneity is poor; the latter have a very good instantaneity and its parameter requirements are not strict, but its adaptability is poor.