基于量测转换与输入估计的机动目标跟踪算法  

Maneuvering target tracking algorithm based on converted measurement and input estimation

作　　者：盛琥[1] 赵温波[1] 王立明[1] 刘发磊[1] 曹燕[1] 

机构地区：[1]陆军军官学院,安徽合肥230031

出　　处：《系统工程与电子技术》2015年第1期31-36,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(61273001);安徽省自然科学基金(1208085QF109)资助课题

摘　　要：针对雷达系统中机动目标的跟踪问题,提出输入估计(input estimation,IE)与无偏量测转换(unbiased converted measurement,UCM)相融合的自适应算法。用无偏量测转换技术将观测量转换到直角坐标系,使得在极坐标系内应用IE成为可能;提出检测窗内加速度阶梯变化的假设并推导相关模型,使得单个输入估计器能够跟踪各种变化机动。仿真分析表明,算法与基于多滤波器并行工作的交互式多模型相比,跟踪精度相当,但运算量大大降低,因此适用于对性能和实时性要求较高的场合,有很好的应用前景。An overall adaptive tracking algorithm is proposed to estimate state variables of maneuvering target. Combining with unbiased converted measurement（UCM）technique to convert polar measurements to Cartesian coordi-nate,input estimation （IE）can be manipulated in polar coordinate.After assuming the acceleration within detection window is linear step distributed,the derivation of relevant model is made,so a single input estimator can track variable maneuver level.Simulation results for different target profiles are included.Compared with the interactive multiple model algorithm which is based on concurrent working of multiple filters,the scheme is comparable in terms of tracking performance while the computation complexity can be significantly reduced,so the proposed algorithm is suitable to ap-plication with high demand for performance and efficiency,which makes it a promising algorithm.

关 键 词：全面自适应跟踪 变加速度跟踪 输入估计 无偏量测转换 

分 类 号：TN953[电子电信—信号与信息处理]