交互多模型PHD滤波跟踪多目标方法

在机动多目标跟踪问题中,目标数未知或随时间而变化,概率假设密度（PHD）滤波可以在每一时间步估计多目标状态和目标数,但单模型方法不能给出精确的估计。提出了一种交互多模型PHD滤波方法,建立多模型描述多目标运动方式,利用PHD滤波结合多模型跟踪目标运动轨迹。同时,给出了多传感器交互多模型PHD滤波方法,以提高目标跟踪精度。     

一种二维耦合模型机动目标跟踪算法

针对雷达跟踪固定翼飞机的单目标跟踪问题,提出了一种新的二维耦合运动模型。该方法把切向加速度和法向加速度作为状态变量,给出了切向加速度和法向加速度为常值时解析计算状态转移的方法。该模型可以较好地一步预测目标加速度的变化,而且法向和切向加速度的过程噪声可以分别设置。利用该耦合模型的滤波方法,显著地改善了滤波效果,尤其是对加速度的估计。     

一种IRST双机协同被动探测机动目标定位新方法

针对红外搜索跟踪系统（IRST）双机协同被动探测定位作战使用中，机动目标建模与实际运动失配造成定位误差偏大的问题，研究了一种基于曲线模型的自适应滤波新方法。该方法改进了传统方法根据方向角估计转弯率以及基于帧间插分线加速度估计切向加速度的思路，将转弯率及线加速度联合作为状态变量进行了状态扩维，并推导了扩维后的过程噪声协方差表达式，在缓解传统两层滤波结构带来的计算量大问题外，也提高了切向加速度的估计精度。另外基于反正切函数的值域，结合方向角在四象限间的转移关系，优化了方向角的设计。通过IRST双机协同仿真实例，验证了所提方法对机动目标的适应性更强、目标定位精度更高。     

混合高斯测量噪声下的多模型滤波

就带有混合高斯测量噪声的离散时间系统 ,提出了一种简化的多模型滤波。理论分析证明该滤波方法用较少的计算量得到了与交互多模型滤波相同的估计性能。为满足应用要求 ,给出了该滤波器的数值鲁棒实现方法。一个关于仅有方位测量的制导例子验证了该滤波算法的有效性。     

基于IMM-UIF的多无人机纯角度机动目标跟踪

针对单无人机不能及时捕捉到目标的运动状态信息,很容易跟丢目标的问题,结合无迹信息滤波(UIF)算法和交互多模型(IMM)算法,提出了基于IMM-UIF的多无人机分布式融合估计算法。将各个无人机上的观测信息传输至中心节点,并统一优化各无人机的控制输入。仿真结果表明,基于IMM-UIF的多无人机分布式融合估计算法比基于IMM-UIF的单无人机跟踪精度提高了约30%,有效融合多无人机平台的量测信息,实现对目标稳定的高精度跟踪。     

图象序列中机动目标的形心跟踪

从边检测边跟踪的角度探讨了图象序列中机动目标的形心跟踪问题,深入分析了强高斯噪声背景下目标形心估计的统计性质及用于形心估计的图象预处理方法。指出经典的图象二值化变换分割后作形心估计的方法面临着估计偏差和方差的矛盾,提出了用自适应交互三模型(ATIMM)跟踪图象序列中机动目标的方法,同时发现在了解目标形状的条件下,空间匹配滤波,二值变换点集聚类和 ATIMM三者的结合对图象序列中的机动目标具有最好的跟踪性能。     

机载雷达目标跟踪与航迹信息提取

针对机动目标跟踪中航迹信息提取精度不高的问题,提出一种ECEF坐标系下基于交互多模型的多机协同跟踪算法。首先,各载机以ECEF坐标系为融合中心对目标量测进行无偏转换处理,以有效减小量测转换误差对目标跟踪的影响;然后,利用交互多模型的方法对目标进行融合跟踪,以进一步提高目标机动时的跟踪精度;最后,通过二次滤波的方法,来有效实现目标航迹信息的精确提取。仿真结果表明,该算法可较好地提高目标机动时的跟踪精度和航迹信息提取精度。     

基于自适应SRCKF的空间非合作目标实时轨道确定

针对天基测角对非合作目标跟踪定轨的动力学模型简化误差问题,提出一种基于非线性预测滤波和SRCKF(Square Root Cubature Kalman Filter,平方根容积Kalman滤波)的自适应滤波方法.采用考虑地球J2摄动影响的轨道动力学模型作为状态方程,在跟踪滤波过程中,用NPF(Nonlinear Predictive Filter,非线性预测滤波)对动力学模型进行实时修正,利用SRCKF对修正后的动力学模型进行状态估计.将该方法应用于高轨航天器对非合作低轨目标的实时测角定轨任务中,进行数字仿真,仿真结果证明,该方法相比传统的滤波方法具有更高的精度、更强的鲁棒性和稳定性.     

常增益变采样率自适应滤波

作者曾给出一种跟踪飞机机动的自适应滤波,采用直角坐标系中的9维状态量,利用最优检测统计量探测目标机动,并用估计到的机动加速度修正目标的预报值。 本文将在文献[1]的基础上,提出两种简化自适应滤波。其一,采用两套直角坐标系中的9维卡尔曼滤波,其中的一套滤波用于取代最优检测统计量,完成对目标机动的探测和估计的功能;其二,采用三通道解偶的常增益变采样率自适应滤波。     

一种基于加速度预估计的机动目标跟踪算法

研究分析了几种典型单机动目标模型的建模方法,针对现有单机动目标模型中机动参数需要先验假设,并且不能随目标机动情况的改变而自适应调整的问题,提出了一种加速度预估计模型(Acceleration Pre-estimation Model,APM)。该模型首先用位置量测对机动加速度进行预估计;然后,将加速度估计值作为系统的输入控制项建模;将估计误差看做系统的机动控制项,并作为系统的相关噪声建模。由于APM模型中,加速度机动参数是通过位置量测实时估计得到的,不需要先验假设。与现有单机动目标模型相比,该模型的自适应能力得到了提高。     

ADAPTIVE MULTIPLE MODEL FILTER USING IMM AND STF

Consider a discrete- time stochastic hybridsystem　 x( k 1 ) =f( k, ( k) ,x( k) ,m( k 1 ) ) ζ( k,m( k 1 ) ) q( k) ( 1 )　 z( k 1 ) =h( k 1 ,x( k 1 ) ,m( k 1 ) ) v( k 1 ,m( k 1 ) ) ( 2 )where state x∈ Rn;measurement z∈ Rm;input∈ Rp;modeling noise q( k)∈ Rqis a zero- mean,Gaussian white noise with covariance Q( k) ;measurement noise v( k 1 )∈ Rm is also a zero-mean,Gaussian white noise with covariance R( k 1 ) ;q( k) and v( k) are statistically indepen-dent. Th…     

引入神经网络的交互式多模型算法

在交互式多模型算法中引入神经网络算法以改进目标跟踪的精度。利用神经网络算法对基于机动目标“当前”统计模型的均值和方差自适应滤波算法进行修改,提高该算法的性能,然后采用交互作用多模型算法跟踪机动目标,提高了机动目标的跟踪精度。     

Efficient fault tolerant estimation using the IMM methodology

Space systems are characterized by a low-intensity process noise resulting from uncertain forces and moments. In many cases, their scalar measurement channels can be assumed to be independent, with one-dimensional internal dynamics. The nominal operation of these systems can be severely damaged by faults in the sensors. A natural method that can be used to yield fault tolerant estimates of such systems is the interacting multiple model (IMM) filtering algorithm, which is known to provide very accurate results. However, having been derived for a general class of systems with switching parameters, the IMM filter does not utilize the independence of the measurement errors in different channels, nor does it exploit the fact that the process noise is of low intensity. Thus, the implementation of the IMM in this case is computationally expensive. A new estimation technique is proposed herein, that explicitly utilizes the aforementioned properties. In the resulting estimation scheme separate measurement channels are handled separately, thus reducing the computational complexity. It is shown that, whereas the IMM complexity is exponential in the number of fault-prone measurements, the complexity of the proposed technique is polynomial. A simulation study involving spacecraft attitude estimation is carried out. This study shows that the proposed technique closely approximates the full-blown IMM algorithm, while requiring only a modest fraction of the computational cost.     

Interacting acceleration compensation algorithm for trackingmaneuvering targets

The two-stage Kalman estimator has been studied for state estimation in the presence of random bias and applied to the tracking of maneuvering targets by treating the target acceleration as a bias vector. Since the target acceleration is considered a bias, the first stage contains a constant velocity motion model and estimates the target position and velocity, while the second stage estimates the target acceleration when a maneuver is detected, the acceleration estimate is used to correct the estimates of the first stage. The interacting acceleration compensation (IAC) algorithm is proposed to overcome the requirement of explicit maneuver detection of the two-stage estimator. The IAC algorithm is viewed as a two-stage estimator having two acceleration models: the zero acceleration of the constant velocity model and a constant acceleration model. The interacting multiple model (IMM) algorithm is used to compute the acceleration estimates that compensate the estimate of the constant velocity filter. Simulation results indicate the tracking performance of the IAC algorithm approaches that of a comparative IMM algorithm while requiring approximately 50% of the computations     

Multisensor tracking of a maneuvering target in clutter

An algorithm is presented for tracking a highly maneuvering target using two different sensors, a radar and an infrared sensor, assumed to operate in a cluttered environment. The nonparametric probabilist data association filter (PDAF) has been adapted for the multisensor (MS) case, yielding the MSPDAF. To accommodate the fact that the target can be highly maneuvering, the interacting multiple model (IMM) approach is used. The results of single-model-based filters and of the IMM/MSPDAF algorithm with two and three models are presented and compared. The IMM has been shown to be able to adapt itself to the type of motion exhibited by the target in the presence of heavy clutter. It yielded high accuracy in the absence of acceleration and kept the target in track during the high acceleration periods     

An interacting multiple model approach for target tracking withglint noise

The application of the interacting multiple model (IMM) estimation approach to the problem of target tracking when the measurements are perturbed by glint noise is considered. The IMM is a very effective approach when the system has discrete uncertainties in the dynamic or measurement model as well as continuous uncertainties. It is shown that this method performs better than the “score function” method. It is also shown that the IMM method performs robustly when the exact prior information of the glint noise is not available     

三维空间中目标机动加速度的估值问题

建立了描述目标在三维空间中进行切向与法向机动的非线性状态模型。目标切向与法向机动加速度的幅值表示为修正的瑞利-马尔可夫随机过程;法向加速度的方向角则假定在2π区间内具有均匀的概率密度。在仅有含噪声位置观察数据的情况下,发展了一种推广的卡尔曼滤波和自适应算法,并由此获得一种机动目标切向与法向加速度估值的直接方法。提供了某些计算结果以证实方法的有效性。     

机动目标“当前”统计模型与自适应跟踪算法

本文提出机动目标“当前”统计模型的概念并建议用修正的瑞利-马尔科夫过程描述目标随机加速机动的统计特性。文中指出了在机动目标运动模型中状态（机动加速度）估值与状态噪声之间的内在联系。在此基础上提出了具有机动加速度均值及方差自适应的卡尔曼滤波算法。对一维和三维的情形进行了计算机模拟。计算结果表明,在仅对目标位置进行观测的情况下,这类自适应估值算法无论对高度机动或无机动的目标均可绘出较好的位置、速度及加速度估值。