Variable structure multiple model fixed-interval smoothing

This paper focuses on fixed-interval smoothing for stochastic hybrid systems.When the truth-mode mismatch is encountered,existing smoothing methods based on fixed structure of model-set have significant performance degradation and are inapplicable.We develop a fixedinterval smoothing method based on forward-and backward-filtering in the Variable Structure Multiple Model(VSMM)framework in this paper.We propose to use the Simplified Equivalent model Interacting Multiple Model(SEIMM)in the forward and the backward filters to handle the difficulty of different mode-sets used in both filters,and design a re-filtering procedure in the model-switching stage to enhance the estimation performance.To improve the computational efficiency,we make the basic model-set adaptive by the Likely-Model Set(LMS)algorithm.It turns out that the smoothing performance is further improved by the LMS due to less competition among models.Simulation results are provided to demonstrate the better performance and the computational efficiency of our proposed smoothing algorithms.

An Algorithm of the Adaptive Grid and Fuzzy Interacting Multiple Model

This paper studies the algorithm of the adaptive grid and fuzzy interacting multiple model （AGFIMM） for maneuvering target tracking, while focusing on the problems of the fixed structure multiple model （FSMM） algorithm＇s cost-efficiency ratio being not high and the Markov transition probability of the interacting multiple model （IMM） algorithm being difficult to determine exactly. This algorithm realizes the adaptive model set by adaptive grid adjustment, and obtains each model matching degree in the model set by fuzzy logic inference. The simulation results show that the AGFIMM algorithm can effectively improve the accuracy and cost-efficiency ratio of the multiple model algorithm, and as a result is suitable for enineering apolications.

Robust estimation algorithm for multiple-structural data

This paper proposes a robust method of parameter estimation and data classification for multiple-structural data based on the linear error in variable（EIV） model.The traditional EIV model fitting problem is analyzed and a robust growing algorithm is developed to extract the underlying linear structure of the observed data.Under the structural density assumption,the C-step technique borrowed from the Rousseeuw＇s robust MCD estimator is used to keep the algorithm robust and the mean-shift algorithm is adopted to ensure a good initialization.To eliminate the model ambiguities of the multiple-structural data,statistical hypotheses tests are used to refine the data classification and improve the accuracy of the model parameter estimation.Experiments show that the efficiency and robustness of the proposed algorithm.

Shipborne radar maneuvering target tracking based on the variable structure adaptive grid interacting multiple model

The trajectory of a shipbome radar target has a certain complexity, randomness, and diversity. Tracking a strong maneuvering target timely, accurately, and effectively is a key technology for a shipbome radar tracking system. Combining a variable structure interacting multiple model with an adaptive grid algorithm, we present a variable structure adaptive grid inter- acting multiple model maneuvering target tracking method. Tracking experiments are performed using the proposed method for five maneuvering targets, including a uniform motion - uniform acceleration motion target, a uniform acceleration motion - uni- form motion target, a serpentine locomotion target, and two variable acceleration motion targets. Experimental results show that the target position, velocity, and acceleration tracking errors for the five typical target trajectories are small. The method has high tracking precision, good stability, and flexible adaptability.

A Class of Distributed Variable Structure Multiple Model Algorithm Based on Posterior Information of Information Matrix

The tracking of maneuvering targets in radar networking scenarios is studied in this paper.For the interacting multiple model algorithm and the expected-mode augmentation algorithm,the fixed base model set leads to a mismatch between the model set and the target motion mode,which causes the reduction on tracking accuracy.An adaptive grid-expected-mode augmentation variable structure multiple model algorithm is proposed.The adaptive grid algorithm based on the turning model is extended to the two-dimensional pattern space to realize the self-adaptation of the model set.Furthermore,combining with the unscented information filtering,and by interacting the measurement information of neighboring radars and iterating information matrix with consistency strategy,a distributed target tracking algorithm based on the posterior information of the information matrix is proposed.For the problem of filtering divergence while target is leaving radar surveillance area,a k-coverage algorithm based on particle swarm optimization is applied to plan the radar motion trajectory for achieving filtering convergence.

一类基于模糊推理的具有机动自适应的目标跟踪算法

针对变结构多模型算法在机动目标跟踪中对目标机动不确定性、量测不确定性自适应能力不足的问题,提出一种基于模糊推理的机动自适应目标跟踪算法.设计一种基于模糊推理的双级机动判别模型,利用模型概率信息和主模型滤波残差加权范数进行主模型可信度和机动判别推理;并将双级机动判别引入基于可能模型集的期望模式扩增方法(EMA-LMS)框架,提出一种模糊推理EMA-LMS算法,实现对模型集自适应的参数和策略的在线调节,从而生成更加接近目标真实运动模式的期望模型,并更好地对模型进行取舍.仿真结果表明,本文算法能够有效增强算法对目标机动和量测不确定的自适应性,提高跟踪精度.

高超声速滑翔飞行器多模型鲁棒跟踪方法

针对高超声速滑翔飞行器(HGV)具有机动能力强、机动样式多变、机动时机不确定等特点,提出了一种基于有向图变结构多模型的鲁棒跟踪(CHF-DSVSMM)方法。考虑到高超声速滑翔飞行器的跳跃滑翔运动特性,建立了包括自适应非零均值衰减震荡(ANMDO)模型、“当前”统计模型(CSM)等机动模型的模型集。针对固定结构多模型算法存在模型相互竞争、运算时间长的缺点,设计了基于有向图结构的切换准则自适应地改变模型集构成,提高模型匹配性。对于地基雷达探测中存在闪烁噪声的问题,采用容积Huber-based滤波方法进行状态估计。仿真结果表明,所提算法与现有方法相比具有更高的跟踪精度,并对闪烁噪声具有良好的鲁棒性。

考虑坐标耦合的三维变结构多模型机动目标跟踪方法

在3维空间机动目标跟踪过程中,目标运动先验未知和坐标耦合误差会引起运动模型-模式失配,而模型-模式失配会引起状态估计有偏。该文根据目标运动速度正交条件修正状态转移矩阵,利用原始-对偶正则约束空间测量到球面可行域,结合自适应转弯率模型和无迹卡尔曼滤波(UKF),进行模型状态滤波并融合状态估计的一致输出,推导3维变结构多模型无迹卡尔曼滤波(VSMMUKF)算法。实验结果表明,相比多模重要性无迹卡尔曼滤波(MIUKF)算法,VSMMUKF计算量相当,能够更准确地拟合3维空间点目标机动运动。相比于交互多模型最大最小粒子滤波(IMM-MPF)算法,VSMMUKF跟踪固定翼无人机(UAV)的滤波精度提升了2.8%~59.9%,整体算法负担减小了1个数量级。

Design and comparison of minimal symmetric model-subset for maneuvering target tracking

Model-set is utilized in state estimation for maneuver- ing target tracking. Two minimal symmetric model-subsets are designed and investigated by moment matching method, which include hypersphere-symmetric model-subset and axis-symmetric model-subset, if system mode is a random variable and obeys certain probability distribution. They can be used as the fun- damental model-subset for multiple models estimation with fixed structure, variable structure and moving bank. The model-groups constructed by above designed subsets are given, which give the practical guidance for use of model-set in multiple models ap- proach with a variable structure. Simulation results show that the performances of two minimal model-set significantly outperform the corresponding model-sets with fixed spacing.

变结构交互式多模型滤波和平滑算法

针对机动目标跟踪问题,提出了一种变结构交互式多模型滤波和平滑算法。首先,对多模型滤波和平滑问题进行了简单描述,并给出了前向交互式多模型滤波和后向交互式多模型平滑的数学模型;然后,建立了变结构交互式多模型算法的精确模型,模型子集之间并行独立运行,通过选取概率最高的模型子集的状态估计作为最终的估计结果;最后,对变结构交互式多模型算法的滤波数据进行平滑处理,得到了变结构交互式多模型滤波和平滑算法。所提算法将前向滤波和后向平滑相结合,提高了目标跟踪精度。仿真结果表明,变结构交互式多模型滤波和平滑算法的跟踪效果优于其他方法。

Model-Based Fault Detection of a Battery System in a Hybrid Electric Vehicle

Recently, a new type of IMM （interacting multiple model） method was introduced based on the relatively new SVSF （smooth variable structure filter）, and is referred to as the IMM-SVSF. The SVSF is a type of sliding mode estimator that is formulated in a predictor-corrector fashion. This strategy keeps the estimated state bounded within a region of the true state trajectory, thus creating a stable and robust estimation process. The IMM method may be utilized for fault detection and diagnosis, and is classified as a model-based method. In this paper, for the purposes of fault detection, the IMM-SVSF is applied through simulation on a simple battery system which is modeled from a hybrid electric vehicle.

一种自适应变结构交互式多模型算法

针对机动目标跟踪问题,提出了一种自适应变结构交互式多模型算法。对滤波问题进行了简单描述,并给出了交互式多模型算法的数学模型;建立了一种新的变结构交互式多模型算法的精确模型,模型子集之间并行独立运行,通过选取概率最高的模型子集的状态估计作为最终的估计结果;在变结构交互式多模型算法基础上进一步设计了自适应变结构交互式多模型算法,该方法提高了模型子集和目标实际机动模式的匹配程度,从而提高了目标跟踪精度。仿真结果表明,自适应变结构交互式多模型算法的跟踪效果优于其他方法。

模型参数自适应的低复杂度ATPM-VSIMM算法

在机动目标跟踪中,针对交互式多模型算法使用固定模型集和固定转移概率矩阵导致跟踪精度下降的问题,提出模型参数自适应更新的低复杂度ATPM-VSIMM算法。所提算法根据系统新息变化情况来判断目标是否出现机动,从而调整模型集的状态噪声,实现模型集的自适应更新;然后,根据模型后验概率变化情况和模型间的相互切换关系,准确地计算出转移概率矩阵,从而提高系统运动模型和目标运动轨迹的匹配程度,保证跟踪系统具有滤波精度高和响应速度快的优点。从模型后验概率初值、转移概率矩阵初值和状态噪声三方面验证了所提算法的有效性。仿真结果表明,ATPM-VSIMM算法的空间位置跟踪精度比现有算法提高了8%左右。

基于结构方程模型的多重中介效应分析

多重中介模型是指存在多个中介变量的模型。多重中介模型可以分析特定中介效应、总的中介效应和对比中介效应。指出了目前多重中介模型分析普遍存在的问题,包括分析不完整、使用Sobel检验带来的局限。建议通过增加辅助变量的方法进行完整的多重中介效应分析,使用偏差校正的Bootstrap方法进行中介检验。总结出一个多重中介SEM分析流程,并有示例和相应的MPLUS和LISREL程序。随后展望了辅助变量和中介效应检验方法的发展方向。

一种跟踪机动目标的变结构交互多模型滤波算法

将当前统计模型和变结构多模型估计算法相结合,阐述了变结构多模型算法的基本思想,研究和设计了一种用于跟踪机动目标的变结构交互多模型算法。计算机仿真显示该算法能有效提高交互多模型估计器的费效比。

变结构多模型临近空间高超声速飞行器跟踪算法

针对传统跟踪算法跟踪临近空间高超声速目标时计算量大、跟踪精度低的问题,提出了一种变结构多模型跟踪算法。在分析传统跟踪算法的基础上,建立了一种变转移概率、变模型数量的目标模型,通过目标运动状态辨识方法,确定目标飞行阶段,并选择相应滤波模型和模型转移概率,实现变结构多模型跟踪。Monte Carlo仿真结果表明:新算法能够实现对临近空间高超声速飞行器的有效跟踪,跟踪性能比传统变结构多模型跟踪算法更加优越,具有一定实用性和可推广性。

基于高机动目标跟踪的改进变结构IMM算法

传统交互式多模型(interacting multiple model,IMM)算法在跟踪高机动目标时存在模型集合和真实系统模式匹配欠佳所导致状态估计质量下降的问题.基于变结构的思想及图论的知识,结合协方差匹配技术提出了一种补偿式变结构交互式多模型算法(compensation variable structure interacting multiple model,CPVSIMM).针对目前对直线加速状态估计问题,传统变结构交互式多模型(variable structure interacting multiple model,VSIMM)模型集多采用角速度作为模型特征参数造成对其估计性能欠佳的情况,所提算法联合加速度和角速度作为模型参数,并建立了模型集合间的有向图连通关系及模型子集自适应调整原则.理论分析与对比仿真表明:本算法融合估计结果的精度更高,同时能在一定程度上减小由于模型集匹配欠佳及切换不及时所导致的机动累积误差.

一种改进的变结构交互多模型被动跟踪算法

针对干扰环境下跟踪机动目标时常常无法获得距离信息的问题,提出了一种新的被动跟踪算法对目标进行精确的跟踪定位。算法针对被动跟踪中状态与量测之间存在的非线性关系,首先采用最小二乘原理对角度量测进行预处理,然后以预处理结果作为输入,进行交互多模型滤波,从而减小非线性量测方程的线性化过程带来的误差。然而,交互多模型滤波中所采用的固定结构模型集并不能保证跟踪精度,为解决这一问题,算法引入序列似然比检测对模型集进行调整,使模型对不同的目标机动模式有更强的自适应能力,从而减小模型之间的竞争,确保跟踪效果。为了验证新算法的有效性,在相同实验条件下,用2种算法分别对同一设定轨迹进行仿真估计,结果表明,新算法可以进一步提高跟踪精度。

机动目标跟踪中的多模算法

机动目标跟踪是一个很具有挑战性的任务 ,即使是跟踪单目标也可能因为目标的逃避机动而失跟。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式存在着不匹配。所以很多学者研究多模算法来解决这一问题。本文对学者们在多模算法上的研究结果进行了系统总结分析 ,得出结论 :变结构的方法是提高跟踪性能、减少运算量 。

机载激光通信终端的模糊变结构跟踪方法研究(英文)

为避免经典固定模型集合中多模型算法造成的模型间竞争并减小计算量,基于有向图切换的变结构方法,建立有向图切换规则和较完备的模型集合.通过模糊逻辑推理得到模型集合中各模型的匹配度,以此代替传统多模型算法中的模式概率来计算,降低了计算的复杂度.采用核粒子滤波克服了标准粒子滤波没有考虑到的量测信息问题,即粒子能够通过核密度估计后朝着状态的后验概率密度的模型移动,使真实和估计模型之间的均方误差最小.仿真结果表明,基于模糊理论的变结构核粒子滤波算法能较大地提高跟踪准确度和减少计算量.