差分进化算法在时差频差高精度实时定位中的应用

Application of Differential Evolution Algorithm on Real-time TDOA and FDOA Passive Localization

通过研究差分进化(DE)算法的基本原理和基于到达时间差(TDOA)和多普勒频差(FDOA)的无源定位的模型,提出了基于差分进化算法的时差频差实时目标定位求解方法和计算流程,同时实现了时间累积的高精度目标估计。仿真表明,即使单次瞬时定位,该定位求解方法都能获得较好的精度,而通过时间累积后精度能接近克拉美劳下界(CRLB),相比于扩展卡尔曼滤波(EKF)算法,收敛更快精度更高。该方法原理简单,易于实现,可以满足工程应用的实时性和精度要求。

By researching on the differential evolution algorithm and passive localization model of two-station TDOA and FDOA, a new calculating method based on DE algorithm is proposed for real-time locating. High precision position estimating employing time cumulate TDOA and FDOA values is also solved by this method. The simulation examples using once and multiple TDOA and FDOA measurements show the more excellent performance on real-time effect and location precision comparing with extended Kalman filter （EKF） algorithm. The location precision almost achieves the Cramer-Rao low bound （CRLB）. Due to the simple principle and easy realization, this method is suitable for project application.