Most actual radar tracking problems are nonlinear ones, in which there are various physical constraints. This paper introduced a method of moving horizon estimation ( MHE) , which was based on the principle of on-line receding optimization, to deal with 9uch nonlinear problems effectively. By introducing the conception of arrival cost and transforming the problem of nonlinear target tracking into the process of constrained optimal state estimation on finite horizon, MHE could effectively reduce the computation burden of solving optimization and notably improve the precision of state estimation. This paper studied the actual nonlinear target tracking problem, and the simulation results show that MHE can effectively improve the tracking accuracy.%实际的雷达跟踪问题大多属于非线性问题,存在着各种物理约束,采用基于在线滚动优化原理的滚动时域估计方法可以有效地处理带约束非线性目标跟踪问题.滚动时域估计通过引入到达代价函数,将非线性跟踪滤波问题转换为带约束的有限时域优化问题,可以有效减少优化问题求解的计算量,能够显著提高状态估计的准确度.针对实际的雷达跟踪问题,仿真结果表明,滚动时域估计能有效地提高非线性目标跟踪的精度.
展开▼
