Unscented Kalman filter(UKF) is widely used in practical engineering, but in the unscented transform(UT) of the filtering process of the traditional UKF, the parameters choose a fixed value, which can bring some errors. In order to obtain the best parameters to be selected, an improved UKF algorithm based on harmony search evolution differential(HSED) is proposed and applied to the target tracking. The harmony differential evolution algorithm can be used to select the optimal parameters, jump out of the local optimum, and has a strong convergence, which can further improve the accuracy of the UKF algorithm. The results of Matlab simulation show that the improved UKF algorithm based on harmonic differential evolution is more accurate.