基于正态截断模型的被动传感器目标跟踪算法

被动传感器只能获得目标的角度信息而无法获得位置信息,因此单被动传感器对目标进行跟踪时难以满足可观测性条件.对单被动传感器高斯-厄米特滤波的量测模型进行扩维,建立了多被动传感器高斯-厄米特滤波模型.由于Singer模型只适用于匀速和匀加速范围内的目标运动,对于强烈的机动将引起较大的模型误差.而正态截断模型本质上是非零均值时间相关模型,能够更加真实地反映目标机动泛围和强度的变化,是目前较好的实用模型.文中基于正态截断模型提出了只有角度量测的双被动传感器高斯-厄密特机动目标跟踪算法,仿真结果表明,该方法能够对机动目标进行稳定的跟踪.

Passive Sensor Target Tracking Algorithm Based on Normal Truncated Model

The passive sensors can only obtain angle information, and cannot obtain the location information of the target. Therefore, the target tracking of a single passive sensor is difficult to meet observability conditions. The thesis focuses on the expansion of the measurement of the single passive sensors Gaussian-Hermitian filtering, and establishes the multiple passive sensors Gaussian-Hermitian filtering model. A larger model error will be caused by the strong motorization due to that the Singer model is only applicable to the target motion within the range of the uniform and uniformly acceleration. Besides, normal truncated model is a better practical model which is essentially the nonzero mean time model, and can more truly reflect changes of motorized range and intensity of the target. The thesis based on normal truncated model proposes the dual passive sensors Gauss-Hermitian maneuvering target tracking algorithm of the angle measurement, and the simulation results show that the method is capable of stably tracking the maneuvering target.