Aiming at problem that in traditional target tracking unscented Kalman filtering（ UKF） algorithm

premise of multi-sensor information fusion is that all sensor observations and dimensions are the same

but it is not suitable for observation systems composed of multiple heterogeneous sensors

an improved UKF filtering algorithm is proposed. In this algorithm

a nonlinear measurement equation in a mixed coordinate system is established by observing fusion information of multiple heterogeneous sensor observations for observation. The simulation results show that this algorithm can effectively reduce target localization error.