适用处理非高斯观测噪声的强跟踪卡尔曼滤波器

在高斯噪声条件下,卡尔曼滤波器(KF)能够获得系统状态的一致最小方差线性无偏估计.但当噪声非高斯,KF性能将严重下降.观测噪声非高斯现象在深空探测自主导航中经常遇到,然而现有模型可能存在着精度不高、稳定性不强或者计算复杂度较高的缺点.针对这种现状,本文在传统强跟踪卡尔曼滤波器(STKF)中新息正交原则的基础上,推导了适用处理非高斯观测噪声的强跟踪卡尔曼滤波器(STKFNO),并将其嵌入到无迹卡尔曼滤波(UKF)框架下形成适用处理非线性系统非高斯观测噪声的强跟踪无迹卡尔曼滤波器(STUKFNO).所提出的算法被应用到深空光学自主导航系统中,仿真结果表明所提出的算法能够较好地应对观测噪声的非高斯性.     

Estimating negative variance components from Gaussian and non-Gaussian data: A mixed models approach

The occurrence of negative variance components is a reasonably well understood phenomenon in the case of linear models for hierarchical data, such as variance-component models in designed experiments or linear mixed models for longitudinal data. In many cases, such negative variance components can be translated as negative within-unit correlations. It is shown that negative variance components, with corresponding negative associations, can occur in hierarchical models for non-Gaussian outcomes as well, such as repeated binary data or counts. While this feature poses no problem for marginal models, in which the mean and correlation functions are modeled directly and separately, the issue is more complicated in, for example, generalized linear mixed models. This owes in part to the non-linear nature of the link function, non-constant residual variance stemming from the mean-variance link, and the resulting lack of closed-form expressions for the marginal correlations. It is established that such negative variance components in generalized linear mixed models can occur in practice and that they can be estimated using standard statistical software. Marginal-correlation functions are derived. Important implications for interpretation and model choice are discussed. Simulations and the analysis of data from a developmental toxicity experiment underscore these results.     

基于高斯和SCKF 的姿态角辅助三维目标跟踪

首先, 根据目标运动与姿态角的关系, 分析目标在偏航角和俯仰角下的速度变化, 进而推导出姿态角辅助三维目标跟踪模型; 然后, 针对姿态角量测非高斯情况, 在分析均方根容积卡尔曼滤波的基础上, 提出新的高斯和均方根容积卡尔曼滤波算法, 以提高非线性非高斯的处理能力; 最后, 结合不同运动模式下姿态角分量的特点, 建立姿态角分量不同的跟踪模型, 通过模型切换实现对姿态角机动的跟踪. 仿真结果验证了所提出跟踪模型和滤波算法的正确性和有效性.     

State estimation of nonlinear dynamic systems using weighted variance-based adaptive particle swarm optimization

New heuristic filters are proposed for state estimation of nonlinear dynamic systems based on particle swarm optimization (PSO) and differential evolution (DE). The methodology converts state estimation problem into dynamic optimization to find the best estimate recursively. In the proposed strategy the particle number is adaptively set based on the weighted variance of the particles. To have a filter with minimal parameter settings, PSO with exponential distribution (PSO-E) is selected in conjunction with jDE to self-adapt the other control parameters. The performance of the proposed adaptive evolutionary algorithms i.e. adaptive PSO-E, adaptive DE and adaptive jDE is studied through a comparative study on a suite of well-known uni- and multi-modal benchmark functions. The results indicate an improved performance of the adaptive algorithms relative to original simple versions. Further, the performance of the proposed heuristic filters generally called adaptive particle swarm filters (APSF) or adaptive differential evolution filters (ADEF) are evaluated using different linear (nonlinear)/Gaussian (non-Gaussian) test systems. Comparison of the results to those of the extended Kalman filter, unscented Kalman filter, and particle filter indicate that the adopted strategy fulfills the essential requirements of accuracy for nonlinear state estimation.     

Blind parametric identification of non-Gaussian FIR systems using higher order cumulants

Two approaches are introduced for the identification of linear time-invariant systems when only output data are available. The input sequences are independent and must be non-Gaussian. To estimate the parameters of the system, we use only the fourth-order cumulants of the output, which may be contaminated by an additive, zero mean, Gaussian noise of unknown variance. To measure the performance of the proposed algorithms against existing methods, we compared them with the Zhang's algorithm. Simulations verify an apparent performance of the second algorithm, compared with the first and Zhang's algorithms, in a low signal-to-noise ratio and for small data. The simulation results show that the first algorithm has the same performance compared with Zhang's one. But the second algorithm achieves better performance compared with the first and Zhang is algorithm. For validation purposes, the second algorithm is used to search for a model able to describe and simulate the data set representing the wind speed.     

Identification of quadratic systems using higher order cumulants and neural networks: Application to model the delay of video-packets transmission

This work concerns the development of two approaches for the identification of diagonal parameters of quadratic systems from only the output observation. The systems considered are excited by an unobservable independent identically distributed (i.i.d), stationary zero mean, non-Gaussian process and corrupted by an additive Gaussian noise. The proposed approaches exploit higher order cumulants (HOC) (fourth order cumulants) and are the extension of the algorithms developed in the linear version 1D, which uses a non-Gaussian signal input. For test and validity purpose, these approaches are compared to recursive least square (RLS), least mean square (LMS) and neural network identification algorithms using non-linear model in noisy environment. To demonstrate the applicability of the theoretical methods on real processes, we applied the developed approaches to search for models able to describe the delay of the video-packets transmission over IP networks from video server. The simulation results show the correctness and the efficiency of the developed approaches.     

Learning linear non-Gaussian networks: A new view from matrix identification

Identifying causal structures from observations is fundamental in many applications. Probabilistic graphical models provide a unifying framework for capturing complex causal dependencies among random variables. Recently, a linear non-Gaussian acyclic model (LiNGAM) and some smart algorithms (ICA-LiNGAM, DirectLiNGAM) have been proposed, which outperform previous graphical models and learning methods in identifying variable orders. We propose new solutions (TMFLiNGAM, SchurLiNGAM and RCLiNGAM) to the LiNGAM learning task from the perspective of matrix identification. TMFLiNGAM and SchurLiNGAM recover orders more directly, and RCLiNGAM can improve the accuracy of previous algorithms on uniform and sparse structures. The perspective also facilitates the learning of sparse models where the performance of all independent component analysis-based algorithms can be improved by reconstructing variable orders from the inverse of separation matrices. Experimental results under various settings provide average evaluations over the learning methods, and verify the effectiveness of our perspective and algorithms.     

Cumulant-based inverse filters for blind deconvolution

Blind deconvolution of linear time-invariant (LTI) systems has received wide attention in various fields such as data communication and image processing. Blind deconvolution is concerned with the estimation of a desired input signal from a given set of measurements. This paper presents a technique for reconstructing the desired input from only the available corrupted data. The estimator is given in terms of an autoregressive moving average (ARMA) innovation model. This technique is based on higher order statistics (HOS) of a non-Gaussian output sequence in the presence of additive Gaussian or non-Gaussian noise. The algorithm solves a set of overdetermined linear equations using third-order cumulants of the given non-Gaussian measurements in the presence of additive Gaussian or non-Gaussian noise. The inverse filter is a finite impulse response (FIR) filter. Simulation results are provided to show the effectiveness of this method and compare it with a recently developed algorithm based on maximizing the magnitude of the kurtosis of estimate of the input excitation.     

ICA mixture models for unsupervised classification of non-Gaussian classes and automatic context switching in blind signal separation

An unsupervised classification algorithm is derived by modeling observed data as a mixture of several mutually exclusive classes that are each described by linear combinations of independent, non-Gaussian densities. The algorithm estimates the density of each class and is able to model class distributions with non-Gaussian structure. The new algorithm can improve classification accuracy compared with standard Gaussian mixture models. When applied to blind source separation in nonstationary environments, the method can switch automatically between classes, which correspond to contexts with different mixing properties. The algorithm can learn efficient codes for images containing both natural scenes and text. This method shows promise for modeling non-Gaussian structure in high-dimensional data and has many potential applications.     

Sea clutter modeling using an autoregressive generalized nonlinear-asymmetric GARCH model

The sea clutter modeling is critical to the radar design and assessment of relevant detection algorithms. In this paper, we investigate a family of generalized autoregressive conditional heteroscedastic (GARCH) processes to model the sea clutter as a time series, in which the current variance is dependent on historical information. The most general model (so-called the ALLGARCH model) provides more flexible variance structures to model non-Gaussian, asymmetry, and nonlinear properties of the clutter. However, after going through the usage of the ALLGARCH model, we find that it is not very suitable because the coefficients of the model, which are numerous, would be difficult to estimate in a real-time operating environment. Meanwhile, we find that some of the coefficients are negligible under almost all kinds of sea environments and weather conditions. Motivated by these observations, we propose a novel GARCH model for sea clutter modeling, which is a generalization of the nonlinear-asymmetric GARCH (NAGARCH) model. Considering the correlation between adjacent clutter returns, autoregressive terms are also introduced. By systematically analyzing practical sea clutter data under different sea environments, we demonstrate that the proposed model achieves comparable fitting effect to some commonly used statistical models. Also, we develop the corresponding generalized likelihood ratio test (GLRT) algorithm for the new model. Numerical simulations exhibit that the proposed detector achieves higher probability of detection, comparing with the AR-GARCH detector.     

基于均差滤波与高斯和的非线性非高斯系统滤波算法

针对一类非线性非高斯系统的滤波问题,在分析均差滤波算法和高斯和滤波算法的基础上,提出一种基于均差滤波的高斯和滤波算法,适于处理非线性非高斯系统的滤波问题.对于似然密度位于条件转移概率密度拖尾处的情况,与传统的粒子滤波算法相比,所提算法能提高滤波的精度和实时性.仿真实验验证了新算法的有效性.     

Classification and recognition of dynamical models: the role of phase, independent components, kernels and optimal transport

We address the problem of performing decision tasks, and in particular classification and recognition, in the space of dynamical models in order to compare time series of data. Motivated by the application of recognition of human motion in image sequences, we consider a class of models that include linear dynamics, both stable and marginally stable (periodic), both minimum and non-minimum phase, driven by non-Gaussian processes. This requires extending existing learning and system identification algorithms to handle periodic modes and nonminimum phase behavior, while taking into account higher-order statistics of the data. Once a model is identified, we define a kernel-based cord distance between models that includes their dynamics, their initial conditions as well as input distribution. This is made possible by a novel kernel defined between two arbitrary (non-Gaussian) distributions, which is computed by efficiently solving an optimal transport problem. We validate our choice of models, inference algorithm, and distance on the tasks of human motion synthesis (sample paths of the learned models), and recognition (nearest-neighbor classification in the computed distance). However, our work can be applied more broadly where one needs to compare historical data while taking into account periodic trends, non-minimum phase behavior, and non-Gaussian input distributions.     

卡尔曼粒子滤波的视频车辆跟踪算法研究

近年来,视频车辆跟踪作为城市智能交通系统（ITS）的一个关键技术受到关注。本文针对传统粒子滤波的非线性、非高斯性可能导致跟踪过程的不稳健性,提出一种基于卡尔曼粒子滤波的视频车辆跟踪算法,该算法利用基于重要区域的目标颜色直方图统计模型对视频车辆目标进行建模,并将其应用于卡尔曼滤波更新中,通过采用Mean Shift算法将卡尔曼滤波器引用到粒子滤波器当中,对车辆的运行轨迹进行校正,实现了局部线性滤波,实现了在保持跟踪系统整体上的非线性、非高斯性的同时,兼顾其局部的线性高斯特性。实验结果表明,本文所提出的方法与传统粒子滤波方法相比,能够更准确地对车辆进行跟踪,同时保证了在复杂环境下性能的稳健性。     

基于多观测模型的粒子滤波头部跟踪算法

粒子滤波在非线性和非高斯问题上具有独特的优越性,但在视频跟踪过程中,其跟踪性能却在很大程度上依赖于观测模型的选择。为了解决被跟踪目标特征状态随时间变化而与粒子观测模型不匹配的问题,提出了一种新的粒子滤波算法,即将被跟踪目标的不同特征状态与粒子观测模型相结合,形成一组具有不同观测模型的粒子,并且在跟踪过程中,对应不同观测模型的粒子根据被跟踪目标所表现的特征线索的变化而相互转换,从而动态刻画了被跟踪目标特征变化的过程。实验结果表明,本算法能够有效处理由于头部旋转而导致跟踪性能下降甚至丢失跟踪目标的问题,提高了跟踪的准确性,并且具有较好的鲁棒性。     

Parameterisation and efficient MCMC estimation of non-Gaussian state space models

The impact of parameterisation on the simulation efficiency of Bayesian Markov chain Monte Carlo (MCMC) algorithms for two non-Gaussian state space models is examined. Specifically, focus is given to particular forms of the stochastic conditional duration (SCD) model and the stochastic volatility (SV) model, with four alternative parameterisations of each model considered. A controlled experiment using simulated data reveals that relationships exist between the simulation efficiency of the MCMC sampler, the magnitudes of the population parameters and the particular parameterisation of the state space model. Results of an empirical analysis of two separate transaction data sets for the SCD model, as well as equity and exchange rate data sets for the SV model, are also reported. Both the simulation and empirical results reveal that substantial gains in simulation efficiency can be obtained from simple reparameterisations of both types of non-Gaussian state space models.     

基于邻域保持的流形学习算法评价模型

应力函数和残差只适合于评价距离严格保持的流形学习算法,dy-dx表示法又是一个定性模型。虽然距离比例方差可以比较和评价大多数的流形学习算法,但其需要计算测地线距离,具有较高的计算复杂度。为此,提出一种基于邻域保持的流形学习算法定量评价模型,该模型仅仅需要确定两个空间中每个对象的k个近邻,并计算出每个点在低维空间中的近邻保持情况,不用计算测地线距离。理论分析表明,邻域保持模型的计算复杂度远远低于距离比例方差的复杂度。在三个数据集上比较了两个模型的性能,实验结果表明,利用邻域保持模型不但可以评价同一算法在不同邻域参数下的嵌入效果,而且可以在不同的流形学习算法之间进行比较,并且其评价流形学习算法的性能优于距离比例方差。     

非高斯噪声下的回声消除算法研究

回声消除一直是信号处理领域的热门研究方向,其中自适应滤波器是在回声消除问题中最为广泛应用的技术,但自适应滤波算法主要是在基于高斯噪声条件下的应用,而现实环境广泛存在着非高斯的噪声,这严重影响了基于L2范数的自适应噪声滤波算法的噪声消除性能。为解决回声消除方法对非高斯噪声的适用性问题,根据回声路径具有明显的稀疏系统特性,结合比例矩阵的设计思想以及符号算法(SA),提出一种改进的MIPNSA算法。该滤波算法既能很好地适应于不同的背景噪声,同时也在较大程度上增强了对稀疏系统的适应能力。仿真测试结果表明,在高斯噪声和非高斯噪声条件下,本算法比现有的一些算法的回声消除效果更佳。     

An evaluation of the nonlinear/non-Gaussian filters for the sequential data assimilation

This paper aims to investigate several new nonlinear/non-Gaussian filters in the context of the sequential data assimilation. The unscented Kalman filter (UKF), the ensemble Kalman filter (EnKF), the sampling importance resampling particle filter (SIR-PF) and the unscented particle filter (UPF) are described in the state-space model framework in the Bayesian filtering background. We first evaluated those methods with a simple highly nonlinear Lorenz model and a scalar nonlinear non-Gaussian model to investigate the filter stability and the error sensitivity, and then their abilities in the one-dimensional estimation of the soil moisture content with the synthetic microwave brightness temperature assimilation experiment in the land surface model VIC-3L. All the results are compared with the EnKF. The advantages and disadvantages of each filter are discussed.The results in the Lorenz model showed that the particle filters are suitable for the large measurement interval assimilation and that the Kalman filters were suitable for the frequent measurement assimilation as well as small measurement uncertainties. The EnKF also showed its feasibility for the non-Gaussian noise. The performance of the SIR-PF was actually not as good as that of the UKF or the EnKF regarding a very small observation noise level compared with the uncertainties in the system. In the one-dimensional brightness temperature assimilation experiment, the UKF, the EnKF and the SIR-PF all proved to be flexible and reliable nonlinear filter algorithms for the low dimensional sequential land data assimilation application. For the high dimensional land surface system that takes the horizontal error correlations into account, the UKF is restricted by its computational demand in the covariance propagation; we must use the EnKF, the SIR-PF and other covariance reduction algorithms. The large computational cost prevents the UPF from being applied in practice.     

Bayesian robust linear dynamic system approach for dynamic process monitoring

In this paper, a Bayesian robust linear dynamic system approach is proposed for process modeling. Traditional linear dynamic system (LDS) constructed with Kalman filter is designed by Gaussian assumption which can be easily violated in non-Gaussian modeling situations, especially those with outliers. To deal with this issue, the conventional Gaussian-based Kalman filter is modified with heavy tailed Student's t-distribution so as to deal with the non-Gaussian noise and modeling outliers. Then, a variational Bayesian expectation maximization (VBEM) algorithm is developed for learning parameters of the robust linear dynamic system. For process monitoring, traditional monitoring scheme are discussed and the residual space monitoring mechanism has been improved. To explore the feasibility and effectiveness, the proposed method is applied for fault detection, with detailed comparative studies with several other methods through the Tennessee Eastman benchmark.     

Robust adaptive spread-spectrum receiver with neural netpreprocessing in non-Gaussian noise

Multiuser communications channels based on code division multiple access (CDMA) technique exhibit non-Gaussian statistics due to the presence of highly structured multiple access interference (MAI) and impulsive ambient noise. Linear adaptive interference suppression techniques are attractive for mitigating MAI under Gaussian noise. However, the Gaussian noise hypothesis has been found inadequate in many wireless channels characterized by impulsive disturbance. Linear finite impulse response (FIR) filters adapted with linear algorithms are limited by their structural formulation as a simple linear combiner with a hyperplanar decision boundary, which are extremely vulnerable to impulsive interference. This raises the issues of devising robust reception algorithms accounting at the design stage the non-Gaussian behavior of the interference. We propose a multiuser receiver that involves an adaptive nonlinear preprocessing front-end based on a multilayer perceptron neural network, which acts as a mechanism to reduce the influence of impulsive noise followed by a postprocessing stage using linear adaptive filters for MAI suppression. Theoretical arguments supported by promising simulation results suggest that the proposed receiver, which combines the relative merits of both nonlinear and linear signal processing, presents an effective approach for joint suppression of MAI and non-Gaussian ambient noise.