基于神经网络补偿的多传感器航迹融合

针对多传感器环境的条件提出了一种基于神经网络补偿的航迹融合方法.各传感器的测量值用线性卡尔曼滤波器进行处理并将获得的局部航迹传送到融合中心.首先对局部航迹进行融合,然后引入神经网络来减少因共同过程噪声而导致的融合估计误差,其中神经网络采用Dan Si-mon提出的网络结构,并对神经网络权值的优化采用无痕卡尔曼滤波(UKF).仿真结果表明,这种融合方法对跟踪具有过程噪声的目标非常有效,而且过程噪声发生变化时该方法仍是有效的,从而使得它在很多实际应用中具有潜在的价值.     

基于交互式多模型扩展维特比容积卡尔曼滤波机场场面运动目标跟踪

根据先进场面运动引导控制系统(advanced surface movement guidance and control system,A-SMGCS)的监视要求,提出一种基于交互式多模型扩展维特比算法(interactive multi-model extended viterbi algorithm,IMMEV)与容积卡尔曼滤波(CKF)相结合的目标跟踪算法,以实现机场场面目标的跟踪。根据机场场面运动目标的运动建立相应的运动模型集,采用交互式多模型算法(IMM)实现目标跟踪;模型集中过多的模型会导致运算复杂性增加,且不能真实反映目标的运动,采用扩展维特比(extended viterbi,EV)算法对IMM进行优化,简化模型集;容积卡尔曼滤波器(cubature Kalman filter,CKF)较传统的滤波算法具有更高的滤波精度和稳定性,将其与IMMEV相结合形成IMMEV-CKF跟踪算法。为了验证所提出的IMMEV-CKF对场面运动目标跟踪性能,仿真研究IMMEV-CKF算法、IMM-UKF和当前统计模型对同一目标的跟踪。仿真结果表明,IMMEVCKF在跟踪精度性能方面要优于IMM-UKF和改进后的当前统计模型。因此,IMMEV-CKF更能满足机场场面运动目标跟踪要求。     

IMM迭代无迹Kalman粒子滤波目标跟踪算法

针对传统交互式多模型(interactive multiple model,IMM)算法跟踪机动式再入目标精度差和实时性不高的问题,提出一种交互式多模型迭代无迹Kalman粒子滤波算法.该算法在多模型滤波过程中采用改进的粒子滤波算法,通过迭代无迹Kalman滤波融入最新观测信息,进而生成粒子滤波的重要性密度分布,从而提高采样质量,改善滤波算法性能.仿真结果表明,提出的算法相对于交互式多模型粒子滤波算法具有更好的跟踪效果.该算法对提高跟踪机动式再入目标的精度与实时能力具有一定的理论意义.     

Cubature Kalman probability hypothesis density filter based on multi-sensor consistency fusion①

The GM-PHD framework as recursion realization of PHD filter is extensively applied to multi-target tracking system .A new idea of improving the estimation precision of time-varying multi-target in non-linear system is proposed due to the advantage of computation efficiency in this paper .First, a novel cubature Kalman probability hypothesis density filter is designed for single sensor measure -ment system under the Gaussian mixture framework .Second , the consistency fusion strategy for multi-sensor measurement is proposed through constructing consistency matrix .Furthermore, to take the advantage of consistency fusion strategy , fused measurement is introduced in the update step of cubature Kalman probability hypothesis density filter to replace the single-sensor measurement .Then a cubature Kalman probability hypothesis density filter based on multi-sensor consistency fusion is proposed .Capabilily of the proposed algorithm is illustrated through simulation scenario of multi-sen-sor multi-target tracking .     

基于空域和时域信息的交互多模优化跟踪算法

机动目标跟踪所遇到的基本问题是所建目标模型的动力学方程与目标的实际运动模型存在着不匹配.为了解决这一问题,提出了基于辐射源信号的方位角及时间差信息的交互多模IMM滤波算法.这种算法是一种最佳的合成滤波器,它的主要特点是具有用几个模型估计系统状态的能力,而这些模型可以相互转换,这种方法能更好地解决对机动辐射源的无源定位跟...     

Maneuvering target tracking algorithm based on cubature Kalman filter with observation iterated update

Reasonable selection and optimization of a filter used in model estimation for a multiple model structure is the key to improve tracking accuracy of maneuvering target.Combining with the cubature Kalman filter with iterated observation update and the interacting multiple model method,a novel interacting multiple model algorithm based on the cubature Kalman filter with observation iterated update is proposed.Firstly,aiming to the structural features of cubature Kalman filter,the cubature Kalman filter with observation iterated update is constructed by the mechanism of iterated observation update.Secondly,the improved cubature Kalman filter is used as the model filter of interacting multiple model,and the stability and reliability of model identification and state estimation are effectively promoted by the optimization of model filtering step.In the simulations,compared with classic improved interacting multiple model algorithms,the theoretical analysis and experimental results show the feasibility and validity of the proposed algorithm.     

禁忌搜索粒子滤波算法在目标跟踪中的应用

雷达目标跟踪量测系统常受到闪烁噪声干扰,导致传统滤波算法的滤波性能急剧下降甚至发散。针对标准粒子滤波算法存在粒子退化的缺陷,重采样环节引入禁忌搜索思想,提出了禁忌搜索扩展卡尔曼粒子滤波算法,驱散局部最优的粒子集,使其向全局最优位置靠近,提高采样粒子的有效性。结合交互多模型（IMM）,将算法与IMM-PF算法进行仿真比较,结果表明该算法对机动目标具有较优的跟踪性能。     

利用径向速度测量值的自适应常增益滤波器

本文推广K.C.Wilson的最优常增盆滤波器到利用径向速度测量值的情况。当目标发生机动时,讨论了机动检测器的设置、速度输入增量的估计和状态预测的修正。本滤波方法与卡尔曼滤波和常增益滤波此较的计算机模拟结果表明:本滤波器减小了机动跟踪时的滤波偏差、提高了机动段内滤波精度,又保持了卡尔曼滤波对非机动目标的良好跟踪性能。     

CDKF方法在外辐射源单站无源目标跟踪中的应用

针对外辐射源单站无源目标跟踪精度问题,将一种基于Sterling内插公式的新的非线性滤波算法—中心差分卡尔曼滤波算法(central difference Kalman filter,CDKF)应用到系统中。仿真结果表明,与传统的扩展卡尔曼滤波算法(extended Kalman filter,EKF)相比,CDKF具有更高的定位精度,且不需要计算Jacobian矩阵,更易实现;而与无迹卡尔曼滤波算法(unscented Kalman filter,UKF)相比,CDKF只有一个待优化的比例参数,参数的选择更加方便,且定位精度更高,所需运算时间更少。     

机动巨标跟踪的自适应混合多模算法

本文介绍了用于机动目标跟踪的自适应混合多模算法.这个算法不需要预先定义模型,它利用一个二级卡尔曼滤波器来估计目标的加速度,这个加速度被用于混合多模算法中具有不同确定性加速度的子滤波器中.文中给出了自适应混合多模算法的一个计算机模拟结果并和无自适应混合多模算法的结果进行了比较.     

“集中式融合的性能一定优于分布式融合的性能”存疑

信息融合领域存在一个公认的“定论”,即集中式融合的性能一定比分布式融合的好,且是最优的;业已证明,当融合系统采用单模型滤波算法,如卡尔曼滤波器时,上述“定论”是成立的;那么在融合系统采用多模型滤波算法,如针对机动目标跟踪的交互多模型滤波(IMM)算法时,情况是否也是如此?抱着对此存疑的态度,对采用IMM算法时的集中式融合与分布式融合进行了研究,给出了扩维、序贯以及等效量测3种集中式融合算法和简单方差凸组合、互协方差组合2种分布式融合算法,并设置不同运动场景对算法进行了大量仿真。仿真实验结果表明,针对交互多模型滤波估计的集中式融合性能并不一定比采用同样滤波算法的分布式融合的好,并给出了作者对此“诧异”现象的认识和理解。     

多传感器粒子滤波融合跟踪算法

对于非线性非高斯环境中的多传感器分布式状态估计问题,提出了一种基于二阶中心差分粒子滤波方法的融合跟踪算法.通过对量测方程的非线性分析,利用粒子滤波器计算目标状态估计值,以在线自适应加权融合算法的方式得到系统最优估计.仿真结果表明,与采用扩展卡尔曼滤波的方法相比,该算法具有更高的估计精度.     

CDKF 算法在四旋翼飞行器组合导航系统中应用

针对基于四旋翼无人飞行器的组合导航系统在小失准角情况下滤波器较短时间内发散的问题, 将中心差分卡尔曼滤波算法应用于低成本的组合导航系统中, 设计了飞行器组合导航系统的硬件电路。该系统采用MPU-6050 芯片作为飞行器姿态传感器, 选用u-Blox 公司的NEO-6M 模块获取GPS(Global Positioning System)信号, 并利用中心差分卡尔曼滤波算法实现多传感器的数据融合, 从而获取导航信息。在四旋翼无人飞行器平台上对该系统进行实验的结果表明, 该导航系统能有效抑制传感器噪声, 克服了陀螺仪的漂移现象, 避免了传统EKF((Extended Calman Filter) )算法近似程度不高, 易导致解算结果发散等问题的发生, 为飞行器提供了准确的导航信息。     

一种机动目标跟踪的IMM模型优化设计方法

为了提高机动目标的跟踪精度,提出了一种使用2个模型实现对机动目标跟踪的多模型算法,采用当前统计模型和扩展后的常速模型组成的模型集进行交互.该算法不受目标转弯率大小和变化的限制,对目标运动模式的未知参数变化的适应性较强.该算法对目标的跟踪精度优于传统的使用3个以上模型交互的IMM-CV/CT算法,仿真结果证明了该算法的有效性.     

Application of interacting multi-model algorithm in gyro signal processing

There is one problem existing in gyroscope signal processing,which is that single models can＇ t adapt to change of carrier maneuvering process.Since it is difficult to identify the angular motion state of gyroscope carriers,interacting multiple model （IMM） is employed here to solve the problem.The Kalman filter-based IMM （IMMKF） algorithm is explained in detail and its application in gyro signal processing is introduced.And with the help of the Singer model,the system model set of gyro outputs is constructed.In order to demonstrate the effectiveness of the proposed approach,static experiment and dynamic experiment are carried out respectively.Simulation analysis results indicate that the IMMKF algorithm is excellent in eliminating gyro drift errors,which could adapt to the change of carrier maneuvering process well.     

光电跟踪系统中高频测角信息的攫取研究

为了能够充分利用光电跟踪设备相邻两次激光测距间隔内的高频测角信息，以提高目标跟踪精度，给出了一种航迹“融合”式目标状态估计方法。该方法借助扩展Kalman滤波算法．将无测距信息时刻的目标测角信息代入独立的滤波器，而对有距离时刻的目标测量信息建立另外的滤波器，然后将两滤波器的输出结果进行加权融合，从而在不增加任何硬件成本的前提下，进一步提高了具有激光测距的光电跟踪系统的目标状态估计精度。对模拟数据和靶场实测数据的测试结果表明了该方法的可行性和有效性。     

基于修正扩展卡尔曼序贯滤波的信息融合算法

针对基于扩展卡尔曼滤波的融合算法存在滤波精度不高的问题,将修正扩展卡尔曼滤波算法与集中式序贯融合算法相结合,用于毫米波雷达和红外传感器目标融合跟踪。即先对毫米波雷达进行修正扩展卡尔曼滤波,再将滤波结果与红外传感器进行融合滤波。仿真结果表明该算法能够提高对机动目标的跟踪精度,增强跟踪系统对环境变化的适应能力。     

A novel maneuvering multi-target tracking algorithm based on multiple model particle filter in clutters

To solve the problem of strong nonlinear and motion model switching of maneuvering target tracking system in clutter environment, a novel maneuvering multi-target tracking algorithm based on multiple model particle filter is presented in this paper. The algorithm realizes dynamic combination of multiple model particle filter and joint probabilistic data association algorithm. The rapid expan- sion of computational complexity, caused by the simple combination of the interacting multiple model algorithm and particle filter is solved by introducing model information into the sampling process of particle state, and the effective validation and utilization of echo is accomplished by the joint proba- bilistic data association algorithm. The concrete steps of the algorithm are given, and the theory a- nalysis and simulation results show the validity of the method.     

基于平方根Unscented卡尔曼滤波的车辆融合跟踪

针对车辆运动的机动性和跟踪系统的非线性，提出了一种基于平方根Unsoented卡尔曼滤波(SR-UKF)的多传感器融合跟踪方法．该方法采用动力学模型建立系统的状态方程和量测方程，充分利用了多传感器的量测信息，更好地满足了目标的机动特性．采用基于UKF的数据融合方法处理系统的非线性问题，避免了扩展卡尔曼滤波(EKF)产生的线性化误差．同时，在滤波过程中，以协方差平方根阵代替协方差阵参加速代运算，有效地避免了滤波器的发散，提高了滤波算法的收敛速度和稳定性，实验证明，与基于EKF的融合算法相比，基于SR-UKF的融合算法使系统的位置和方向角的跟踪精度分别提高了18．22％和34．81％。     

Federal extended Kalman filter based on reconstructed observation in incomplete observations

In the estimation and identification of nonlinear system state,aiming at the adverse effect of observation missing randomly caused by detection probability of used sensor which is less than 1,a novel federal extended Kalman filter( FEKF) based on reconstructed observation in incomplete observations( ROIO) is proposed in this paper. On the basis of multi-sensor observation sets,the observation is exchanged at different times to construct a new observation set. Based on each observation set,an extended Kalman filter algorithm is used to estimate the state of the target,and then the federal filtering algorithm is used to solve the state estimation based on the multi-sensor observation data. The effect of the sensor probing probability on the filtering result and the effect of the number of sensors on the filtering result are obtained by the simulation experiment,respectively. The simulation results demonstrate effectiveness of the proposed algorithm.