Abstract: An “inverse” unscented Kalman filter (UKF) algorithm with seven-dimensional state vector is proposed to solve the problems of low target positioning accuracy and poor fire direction ability of firefin-der radar. A state model is established by taking the ballistic coefficient as the state parameter and incorporating it into the filtering process. The UKF algorithm is used to improve the nonlinear estimation accuracy. The model error accumulates due to the long extrapolated distance of forward filtering. In the proposed algorithm, an inverse filtering is used, the first point measured by radar is used as the end point of the filter, and the artillery position is extrapolated using the fourth-order Runge-Kutta equation. The simulation results show that the proposed algorithm can effectively improve the extrapolation accuracy of artillery locating and fire correction radar. Key

Key words: unscentedKalmanfilter, ballisticcoefficient, inversefiltering, Runge-Kuttaequation