论文部分内容阅读
为了解决单红外观测站对目标进行跟踪时存在的非线性估计问题,提出了基于直角坐标系的多假设高斯-厄密特滤波算法.为了减小距离不可观测性对于非线性滤波的影响,在假设的多段距离间隔中采用并行的高斯-厄密特滤波,并加权获得目标的状态估计.首先将初始时刻红外观测站的观测距离区间划分为若干个子区间,每个子区间表示关于目标真实距离的一种假设;在每个子区间内用共同的量测进行独立的高斯-厄密特滤波,并将每个子区间对应的概率根据贝叶斯规则进行递归计算;最后将各子区间的状态估计和协方差进行加权求和得到最终估计.仿真实验结果表明,由于使用高斯-厄密特滤波从而避免求解雅克比矩阵,且距离多假设降低了不确定性,故该算法的滤波精度高于高斯-厄密特滤波和扩展卡尔曼滤波.
In order to solve the problem of non-linear estimation when tracking a target with a single infrared observatory, a multi-hypothesis Gaussian-Hermitian filter based on Cartesian coordinate system is proposed.In order to reduce the effect of distance unobservability on nonlinear filtering, Gaussian-Hermitian filter with parallel Gaussian-Hermite filter is adopted in the assumed multi-segment distance interval, and the state estimation of the target is obtained by weighting.Firstly, the observation distance interval of the infrared observatory at the initial moment is divided into several subsections, each representing the true distance One hypothesis: Independent gauss-Hermitian filtering with common measurements in each subinterval, and recursively calculate the corresponding probabilities of each subinterval according to Bayes rule. Finally, the state estimates of each subinterval and And the covariance is weighted by the summation to get the final estimate.The simulation results show that the filtering accuracy of the proposed algorithm is higher than that of Gaussian-Hermitian filter because it avoids solving the Jacobian matrix and the distance hypothesis reduces the uncertainty, Hermitian and Extended Kalman Filtering.