根据2D传感器的扫描特点和观测信息,建立了单站2D传感器对三维运动目标的观测模型,基于该模型提出了三维 匀速运动目标的仰角参数化扩展卡尔曼滤波算法。此算法将传感器的仰角观测范围划分为若干个子区间,对各仰角子区 间执行并行独立的EKF跟踪算法,并根据滤波结果动态更新各滤波器的权值,最后融合各EKF估计值得到最终的目标状 态估计。计算机仿真结果证明了该算法在2D传感器对三维空间内目标跟踪定位时的有效性和优越性。

According to the scan characteristic and observation information of 2D sensor, the observation model for three-dimensional moving target using single station was built, and an elevation parameterized extended kalman filter (EPEKF) algorithm for target in three-dimensional constant velocity was presented. Firstly, the elevation region of the sensor was divided into several subintervals. Then, the algorithm performs parallel absolute extended Kalman filter in each subinterval and the weights of each filter was updated according to the tracking result. Finally, the estimation from each filter was fused to obtain the final state estimation of the target. The accuracy and stability of the proposed tracking and location algorithm for the target in 3D space can be confirmed by the simulation results.

国家自然科学基金