检索结果-内蒙古大学图书馆

robust sparse representation based on fitting error decomposition

SIGNAL PROCESSING 2024年 222卷

作者： Wang, Xiang -Yu Li, Xiao-Peng So, Hing Cheung City Univ Hong Kong Dept Elect Engn Hong Kong Peoples R China Shenzhen Univ Coll Elect & Informat Engn Shenzhen 518060 Peoples R China

Sparse representation (SR) of a signal aims at finding the minimum number of atoms for its representation. In several practical scenarios, the signal is vulnerable to outliers and thus robustness is required for SR based algorithms. However, most existing robust schemes are designed on the assumption that the anomalies are independently distributed, which may not perform well encountering complex interferences, especially the correlated ones. To deal with this drawback, a robust SR model is proposed, where both independent and correlated outliers are considered. Specifically, the fitting error is decomposed as the combination of a low -rank component and a sparse part, corresponding to the correlated gross error and independent outlier, respectively. Then, the group sparsity of the representation coefficient is utilized. Moreover, t degrees 0 -norm and t degrees 2 , 0 -norm are adopted as the sparsity regularization for the sparse outlier and representation coefficients, respectively. The solutions to t degrees 0 /t degrees 2 , 0 -norm minimization are generated by the hard-thresholding strategy, where the decision threshold is adaptively determined using median absolute deviation operator. As for the low -rank regularization, due to the NP -hardness of rank minimization, we employ matrix logarithmic norm as the rank surrogate to lessen the approximation gap. Finally, we apply the proposed model to face recognition task, and the excellent performance demonstrates its effectiveness.

关键词： robust algorithm Sparse representation Outlier t degrees 0-norm Logarithmic norm

来源：评论

学校读者我要写书评

暂无评论

Trimmed-Mean FxLMS algorithm for ANC in Impulsive Noise 10

Trimmed-Mean FxLMS Algorithm for ANC in Impulsive Noise

引用

10th Asia-Pacific-Signal-and-Information-Processing-Association Annual Summit and Conference (APSIPA ASC)

作者： Chang, Dah-Chung Chiang, Po-Wei Natl Cent Univ Dept Commun Engn Taoyuang 32001 Taiwan

ISBN: (纸本)9789881476852

The filtered-x least mean square ( FxLMS) algorithm has been widely used for active noise control ( ANC), which has a satisfying convergence performance in the Gaussian noise. Nevertheless, the ANC system can receive an unusual signal contaminated by the impulsive noise ( IN) in practical applications, which results in serious convergence performance degradation. Some references can be found discussing this issue, but many of the proposed solutions highly rely on the setup of pre-determined parameters depending on special situations. In this paper, the IN problem is dealt with by some robust algorithms, which are adapted from order statistics. With the new FxLMS algorithms, there is no pre-requisite to determine the parameters as required in the past works. Through simulation results, we show that the proposed method has better average noise reduction convergence performance compared to other robust methods.

关键词： Active noise control Impulsive noise FxLMS Order statistics robust algorithm

来源：评论

学校读者我要写书评

暂无评论

Ensuring the robustness of the algorithm for Detecting Signals of Moving Targets against the Background of Non-Gaussian Clutter 2

Ensuring the Robustness of the Algorithm for Detecting Signa...

引用

IEEE 2nd Ukrainian Microwave Week (UkrMW)

作者： Prokopenko, Igor Dmytruk, Anastasiia Natl Aviat Univ Dept Telecommun & Radioelect Syst Kiev Ukraine

ISBN: (纸本)9798350331523

In this paper the problem of ensuring the stability of the algorithm for detecting radar signals reflected from a moving target against the background of non-Gaussian interference is considered. An empirical Bayesian approach for the synthesis of decision-making rules is implemented using an autoregressive disturbance model, and a robust adaptive algorithm is also developed. Efficiency of the synthesized algorithm under the influence of non-Gaussian interference distributed according to the laws of Laplace and K-distribution is defined. Computer simulation was carried out using the Monte Carlo method.

关键词： autoregressive process robust algorithm clutter moving target indication iteration algorithm

来源：评论

学校读者我要写书评

暂无评论

Partitioned Latency Insertion Method (PLIM) with Stability Considerations

Partitioned Latency Insertion Method (PLIM) with Stability C...

引用

15th IEEE Workshop on Signal Propagation on Interconnects (SPI)

作者： Goh, Patrick Schutt-Aine, Jose E. Klokotov, Dmitri Tan, Jilin Liu, Ping Dai, Wenliang Al-Hawari, Feras Univ Illinois Dept Elect & Comp Engn Champaign IL 61820 USA Cadence Design Syst Chelmsford MA 01824 USA

ISBN: (纸本)9781457704673

In this paper, we present a modular approach to the high frequency simulation of large networks by utilizing the latency insertion method (LIM) and considering the stability criteria of partitions of different latencies in the circuit. This results in a robust algorithm that is able to preserve the stability condition while improving the runtime of the overall transient simulation. An extension to the LIM to handle dependent sources is also presented, along with a generalized stability criteria for selecting a maximum stable time step.

关键词： Equations Integrated circuit modeling Niobium RLC circuits Stability criteria Topology circuit simulation circuit stability frequency simulation integrated circuit interconnections large networks partitioned latency insertion method robust algorithm stability condition transient simulation

来源：评论

学校读者我要写书评

暂无评论

Multi-Stage Stabilization of ROMs in Strongly Nonlinear Systems

Multi-Stage Stabilization of ROMs in Strongly Nonlinear Syst...

引用

AIAA Aviation Forum

作者： Rezaian, Elnaz Wei, Mingjun Kansas State Univ Dept Mech & Nucl Engn Manhattan KS 66506 USA

ISBN: (数字)9781624105982

ISBN: (纸本)9781624105982

The standard model order reduction procedure by POD-Galerkin projection does not guarantee the closure of reduced-order dynamical systems, thus the stability and accuracy of the models may be compromised when the dynamically observable signatures are not populated in low-frequency responses. Turbulence and unsteady shock waves, among an abundance of fluid mechanics phenomena, are two examples with important high-frequency or small-scale contributions. Efficient stabilization algorithms are therefore essential to the projection-based model reduction procedure. When strong unsteady shock-wake interactions, a highly nonlinear phenomenon, are present in the flow, a linearized Reduced-Order Model (ROM) is not accurate even after stabilization. A nonlinear ROM on the other hand may keep the accuracy after stabilization, but requires a fairly large number of bases for convergence. Among a few stabilization methods previously developed for nonlinear ROMs, the ERN algorithm is used in this study to stabilize the ROM by an eigenvalue reassignment method implemented as an optimization problem constrained by negative total power. Despite its computational efficiency in smaller ROMs, the ERN algorithm becomes computationally more expensive and may fail as the number of unstable modes increases in ROMs with larger number of bases. To solve this problem in the stabilization of relatively large-size nonlinear ROMs, a multi-stage approach is proposed to divide the optimization process of the ERN algorithm into a few stages, where each stage contains only a small number of unstable modes for an overall more efficient and robust algorithm. It is shown that the computational cost and accuracy of the ERN algorithm significantly benefits from the new stabilization layout in the application to a nonlinear ROM created from the snapshots of the supersonic flow over a triangular prism.

关键词： Nonlinear Systems Compressible Flow Model Order Reduction Shock Waves Frequency Response robust algorithm Fluid Mechanics Model Predictive Control Euler Equations Direct Numerical Simulation

来源：评论

学校读者我要写书评

暂无评论

The LFOPC leap-frog algorithm for constrained optimization

引用

COMPUTERS & MATHEMATICS WITH APPLICATIONS 2000年第8-9期40卷 1085-1096页

作者： Snyman, JA Univ Pretoria Dept Mech & Aeronaut Engn Multidisciplinary Design Optimizat Grp ZA-0002 Pretoria South Africa

This paper describes an accurate and reliable new algorithm (LFOPC) for solving constrained optimization problems, through a three-phase application of the well-established leap-frog method for unconstrained optimization, to penalty function formulations of the original constrained problems. The algorithm represents a considerable improvement over an earlier version (LFOPCON) which requires the judicious choice of parameter settings for efficient use. The current algorithm automatically executes normalization and scaling operations on the gradients of the constraints. This results in a robust algorithm that, apart from convergence tolerances, requires virtually no parameter settings. The method has been well tested, on both standard analytical test problems and practical engineering design problems. (C) 2000 Elsevier Science Ltd. All rights reserved.

关键词： constrained optimization leap-frog method robust algorithm penalty function formulations

来源：评论

学校读者我要写书评

暂无评论

New Polynomial Cases of the Weighted Efficient Domination Problem

New Polynomial Cases of the Weighted Efficient Domination Pr...

引用

38th International Symposium on Mathematical Foundations of Computer Science (MFCS)

作者： Brandstaedt, Andreas Milanic, Martin Nevries, Ragnar Univ Rostock Inst Informat D-18051 Rostock Germany Univ Primorska UP IAM SI-6000 Koper Slovenia Univ Primorska UP FAMNIT SI-6000 Koper Slovenia

ISBN: (纸本)9783642403132;9783642403125

Let G be a finite undirected graph. A vertex dominates itself and its neighbors in G. A vertex set D is an efficient dominating set (e.d. for short) of G if every vertex of G is dominated by exactly one vertex of D. The Efficient Domination (ED) problem, which asks for the existence of an e.d. in G, is known to be NP-complete even for very restricted graph classes. In particular, the ED problem remains NP-complete for 2P(3)-free graphs and thus for P-7-free graphs. We show that the weighted version of the problem (abbreviated WED) is solvable in polynomial time on various subclasses of P-7-free graphs, including (P-2 + P-4)-free graphs, P-5-free graphs and other classes. Furthermore, we show that a minimum weight e.d. consisting only of vertices of degree at most 2 (if one exists) can be found in polynomial time. This contrasts with our NP-completeness result for the ED problem on planar bipartite graphs with maximum degree 3.

关键词： efficient domination P-k-free graphs polynomial-time algorithm robust algorithm

来源：评论

学校读者我要写书评

暂无评论

A novel framework for unstructured meshes with optimized cell ordering using Hamiltonian paths

A novel framework for unstructured meshes with optimized cel...

引用

AIAA Science and Technology Forum and Exposition (AIAA SciTech Forum)

作者： Amouyal, Solal A. T. Margolin, Alex Toga Networks Tel Aviv Israel

ISBN: (数字)9781624106095

ISBN: (纸本)9781624106095

A novel optimized framework for unstructured mesh is presented, in which cells are ordered using Hamiltonian paths to maximize data locality. Edge-based Hamiltonian paths have the advantage of guaranteeing that two consecutive cells in the path are contiguous in memory, greatly reducing the number of failed data requests (i.e. cache-misses) inherent to unstructured grids. Interior and boundary cells are treated separately to take full advantage of advanced asynchronous MPI communication and reduce the time exchanging ghost cells. To further increase data locality, multiple levels of mesh partitioning are used, each resulting submesh having its own Hamiltonian paths. As a preliminary study, a cell-centered finite-volume Poisson solver using the Jacobi method was implemented using the original and Hamiltonian path cell ordering on meshes of varying sizes. Serial simulations show a speed-up of up to 1.5 and 2.6 for the full Poisson solver and the corresponding cell-traversing algorithm, respectively, when compared to the original cell ordering mesh. Furthermore, profiling results showthat the last-level cache loads and misseswere reduced by up to76% and 78%, respectively, demonstrating the increase in data locality and overall performance gains of the Hamiltonian path cell ordering.

关键词： Unstructured Grid Jacobi Method Computational Fluid Dynamics Adaptive Mesh Refinement Industrial Applications Asynchronous Communications robust algorithm Turbulent Flow Supercomputers

来源：评论

学校读者我要写书评

暂无评论

Fourier Subspace Preserving Particle Filter (Poster) 22

Fourier Subspace Preserving Particle Filter (Poster)

引用

22nd International Conference on Information Fusion (FUSION)

作者： Thron, Christopher Tran, Khoi Texas A&M Univ Cent Texas Dept Sci & Math Killeen TX 76549 USA

ISBN: (纸本)9780996452786

We introduce a new particle filter, based on the mathematical notion of weak approximation. The filter is similar to the Daum-Huang particle filter in that it propagates filter particles based on a "homotopy" that is used to adjust the position of particles. However, the homotopy that we propose is much simpler, and does not require extraneous mathematical constructions. We verify the filter's performance in one and two dimensions. The filter holds promise of providing a low complexity, high performance filter.

关键词： filtering particle filters robust algorithm

来源：评论

学校读者我要写书评

暂无评论

Multilayer Tensioned Membrane Structures for Radio -Frequency Lenses

Multilayer Tensioned Membrane Structures for Radio -Frequenc...

引用

AIAA SciTech Forum

作者： Park, Jung Eun Brownt, Garrett C. Arya, Manan Hoppe, Daniel Hofmann, Douglas Hodges, Richard Stanford Univ Dept Aeronaut & Astronaut 496 Lomita Mall Stanford CA 94305 USA CALTECH Jet Prop Lab Pasadena CA 91125 USA

ISBN: (数字)9781624107115

ISBN: (纸本)9781624107115

This study explores the structural implementation of a multilayer tensioned membrane structure to build a transmission -based metalens radio frequency (RF) aperture. Its advantages include relaxation in mechanical tolerances, lightweight, and compact stowage. The structure is realized with lOs of layers of tensioned membrane with sub -millimeter layer separation. Discrete spacers are used to maintain the sub -millimeter separation along with tensioning mechanisms that applies equal tension to all layers for equal stiffness. The study validates repeatability and tension uniformity of early stage prototypes. Using experiments to measure the shape of relevant prototypes with digital image correlation (DIC), we confirm that the proposed design is highly repeatable within the range to tens of microns. We demonstrate prototype tensioning mechanisms that achieve equal tension within 0.9% over 4 layers of membrane.

关键词： Membrane Structures Finite Element Modeling Synthetic Aperture Radar Thermoplastic Polyurethane MATLAB Trusses Structural Design Thermal Stability robust algorithm Tear Resistance

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：