检索结果-内蒙古大学图书馆

第41届中国控制会议

作者： Jianhao Li Xiaowei Jiang Hao Chen Shaoying Wang Xianhe Zhang School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation of Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Fujian Provincial Key Laboratory of Intelligent Identification and Control of Complex Dynamic System College of Science Binzhou University School of Electrical Engineering and Automation Hubei Normal University

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

In this paper,we focus on the tracking performance(TP) limitations of networked control systems(NCS s).The control system considered is a multi-input and multi-output(MIMO) discrete-time network control system with time delay,channel noise,packet dropouts,and logarithmic quantization communication *** modeling the channel,we assumed the channel noise to be additive white Gaussian noise(AWGN) and used a binary random process to model packet dropouts and a logarithmic quantizer to model the quantization *** explicit expression of the TP limitations of NCSs utilizing the two-degree-of-freedom(TDOF) controller is then derived using the frequency-domain approach,coprime decomposition,and Youla parameterization *** results show that the essential characteristics of the plant(locations and directions of nonminimum phase zeros and unstable poles) and network constraints(time delay,packet dropouts,quantization,and noise power spectral density) are closely related to the TP limitations.

关键词： Networked control systems tracking performance limitations packet dropouts logarithmic quantization

来源：评论

学校读者我要写书评

暂无评论

Jumping Shift: A logarithmic quantization Method for Low-Power CNN Acceleration

Jumping Shift: A Logarithmic Quantization Method for Low-Pow...

引用

Design, Automation and Test in Europe Conference and Exhibition (DATE)

作者： Jiang, Longxing Aledo, David Van Leuken, Rene Delft Univ Technol Circuits & Syst Grp Delft Netherlands

ISBN: (纸本)9798350396249

logarithmic quantization for Convolutional Neural Networks (CNN): a) fits well typical weights and activation distributions, and b) allows the replacement of the multiplication operation by a shift operation that can be implemented with fewer hardware resources. We propose a new quantization method named Jumping Log quantization (JLQ). The key idea of JLQ is to extend the quantization range, by adding a coefficient parameter "s" in the power of two exponents (2(sx+i)). This quantization strategy skips some values from the standard logarithmic quantization. In addition, we also develop a small hardware-friendly optimization called weight de-zero. Zero-valued weights that cannot be performed by a single shift operation are all replaced with logarithmic weights to reduce hardware resources with almost no accuracy loss. To implement the Multiply-And-Accumulate (MAC) operation (needed to compute convolutions) when the weights are JLQ-ed and de-zeroed, a new Processing Element (PE) have been developed. This new PE uses a modified barrel shifter that can efficiently avoid the skipped values. Resource utilization, area, and power consumption of the new PE standing alone are reported. We have found that JLQ performs better than other state-of-theart logarithmic quantization methods when the bit width of the operands becomes very small.

关键词： Convolutional Neural Network Low-power hardware acceleration logarithmic quantization FPGA

来源：评论

学校读者我要写书评

暂无评论

RATIONAL DITHER MODULATION USING logarithmic quantization WITH OPTIMUM PARAMETER

RATIONAL DITHER MODULATION USING LOGARITHMIC QUANTIZATION WI...

引用

IEEE International Conference on Acoustics, Speech, and Signal Processing

作者： Nima Khademi Kalantari Seyed Mohammad Ahadi Electrical Engineering Department Amirkabir University of Technology 424 Hafez Avenue Tehran 15914 Iran

ISBN: (纸本)9781424442959

In this paper, a new logarithmic quantization for Rational Dither Modulation (RDM) is presented. It can be shown that the mu-law function produces quantization levels which are in best accordance with the noise characteristics in RDM. However, since the mu-law function cannot be used in its original form, as in [5], we use it with slight modifications. In order to obtain the optimum quantizer arrangement, which results in minimum error probability, the analytical error probability is obtained and then minimized with respect to mu, which defines the compression level of the logarithmic function. The embedding distortion for the proposed method is also derived. Simulation results show that, using logarithmic quantization, better performance can be achieved in comparison with conventional RDM. Furthermore, as will be shown, the proposed method is more suitable in practice due to its perceptual advantages.

关键词： quantization Index Modulation (QIM) Rational Dither Modulation (RDM) logarithmic quantization digital watermarking Digital watermarking noise characteristics quantization logarithmic Error probability quantization level ORIGINAL

来源：评论

学校读者我要写书评

暂无评论

Asymptotic Stabilization for Uncertain Nonlinear Systems With Input quantization

引用

IET CONTROL THEORY AND APPLICATIONS 2025年第1期19卷

作者： Yan, Fei Wang, Shuo Gu, Guoxiang Southwest Jiaotong Univ Sch Informat Sci & Technol Chengdu Peoples R China Louisiana State Univ Sch Elect Engn & Comp Sci Baton Rouge LA 70803 USA

This paper investigates the problem of asymptotic stabilization for a class of uncertain nonlinear systems involving logarithmic quantization at the system input. Different from the existing results and approaches, a Lyapunov function candidate and an adaptive control law are developed to adaptively estimate the uncertain parameters and to asymptotically stabilize the uncertain nonlinear system, in which the control input also involves uncertain parameters, possibly in the nonlinear form. It is shown that asymptotic stabilization can be achieved under some mild conditions, even though the adaptively estimated parameters do not converge to the true system parameters. A sufficient condition is obtained for the asymptotic stabilizability, in terms of the quantization density and the multiplicative parameter error bound at the control input. More importantly, the proposed adaptive control law is suboptimal for the corresponding LQR control and achieves the H-infinity-norm to be strictly smaller than gamma, provided that gamma>1, for the uncertain linearized closed-loop system, effectively suppressing energy bounded disturbances. Finally, two simulation examples are worked out to illustrate the effectiveness of the proposed method.

关键词： adaptive parameter estimation asymptotic stabilization logarithmic quantization uncertain nonlinear systems

来源：评论

学校读者我要写书评

暂无评论

Further Results on the R-R Protocol Based H∞ Estimator for 2-D FMII Systems with R-O Uncertainties

引用

CIRCUITS SYSTEMS AND SIGNAL PROCESSING 2025年第4期44卷 2424-2453页

作者： Qobbi, Hicham Zoulagh, Taha Boukili, Bensalem Barbosa, Karina A. Hmamed, Abdelaziz Chaibi, Noreddine Sidi Mohamed Ben Abdellah Univ Fez LISAC Lab Fes Morocco Univ Santiago Santiago Chile Private Univ Fez Fes Morocco

This paper presents an enhanced study on the design of an H-infinity observer based on the Round-Robin protocol for 2-D FMII systems subjected to Randomly-Occurring uncertainties. The Randomly-Occurring uncertainties, characterized by their stochastic nature and random occurrence, pose significant challenges to the stability and performance of control systems. Moreover, the Round-Robin protocol, known for its efficiency in managing resources and reducing data conflicts in communication systems, is used as the framework for developing this estimator. The main contributions of this research include a precise modeling of Randomly-Occurring uncertainties and their integration into the 2-D FMII system model. A robust H-infinity estimator is designed ensuring optimal performance despite the presence of stochastic uncertainties, and the application of the Round-Robin protocol to synchronize estimation and control processes, thereby enhancing system resilience. On the basis of the Projection Lemma, less conservative results in terms of Linear Matrix Inequalities are provided. The theoretical results are validated and compared through numerical simulations from the literature, demonstrating the effectiveness and robustness of the proposed approach.

关键词： H-infinity estimation RRP logarithmic quantization Randomly occurring (R-O)uncertainties 2-D systems

来源：评论

学校读者我要写书评

暂无评论

MXQN:Mixed quantization for reducing bit-width of weights and activations in deep convolutional neural networks

引用

APPLIED INTELLIGENCE 2021年第7期51卷 4561-4574页

作者： Huang, Chenglong Liu, Puguang Fang, Liang Natl Univ Def Technol Coll Comp Sci & Technol Changsha 410073 Peoples R China

quantization, which involves bit-width reduction, is considered as one of the most effective approaches to rapidly and energy-efficiently deploy deep convolutional neural networks (DCNNs) on resource-constrained embedded hardware. However, bit-width reduction on the weights and activations of DCNNs seriously degrades accuracy. To solve this problem, in this paper we propose a mixed hardware-friendly quantization (MXQN) method that applies fixed-point quantization and logarithmic quantization for DCNNs without the necessity to retrain and fine-tune the DCNN. Our MXQN algorithm is a multi-staged process where, first, we employ a signal-to-quantization-noise ratio (SQNR) process as the metric to estimate the interplay between the parameter quantization errors of each layer and the overall model prediction accuracy. Then, we utilize a fixed-point quantization process to quantize weights, and depending on the SQNR metric we empirically select either a logarithmic or a fixed-point quantization process to quantize activations. For improved accuracy, we propose an optimized logarithmic quantization scheme that affords a fine-grained step size. We evaluate the performance of MXQN utilizing the VGG16 network on the MNIST, CIFAR-10, CIFAR-100, and the ImageNet datasets, as well as VGG19 and ResNet (ResNet18, ResNet34, ResNet50) networks on the ImageNet, and demonstrate that the MXQN-quantized DCNN despite not being retrained and fine-tuned, it still achieves high accuracy close to the original DCNN.

关键词： Mixed quantization logarithmic quantization Fixed-point quantization Signal-to-quantization-noise ratio (SQNR) No retraining Resource-constrained

来源：评论

学校读者我要写书评

暂无评论

Non-fragile Suboptimal Set-membership Estimation for Delayed Memristive Neural Networks with quantization via Maximum-error-first Protocol

引用

INTERNATIONAL JOURNAL OF CONTROL AUTOMATION AND SYSTEMS 2020年第7期18卷 1904-1914页

作者： Yang, Yu Hu, Jun Chen, Dongyan Wei, Yunliang Du, Junhua Harbin Univ Sci & Technol Dept Math Harbin 150080 Peoples R China Harbin Univ Sci & Technol Heilongjiang Prov Key Lab Optimizat Control & Int Harbin 150080 Peoples R China Univ South Wales Sch Engn Pontypridd CF37 1DL M Glam Wales Qufu Normal Univ Sch Math Sci Qufu 273165 Shandong Peoples R China Qiqihar Univ Coll Sci Qiqihar 161006 Peoples R China

This paper is concerned with the non-fragile protocol-based set-membership estimation problem for a class of discrete memristive neural networks (MNNs) with mixed time-delays, quantization and unknown but bounded noises. The nonlinear neural activation function satisfies the sector-bounded condition and the logarithmic quantization error is transformed to the norm-bounded uncertainty. In order to save the networks resources, the maximum-error-first (MEF) protocol is introduced to allocate the utilization order of the network channel. The focus is on the design of non-fragile state estimator to ensure such that, in the simultaneous presence of the mixed time-delays, quantization errors and estimator gain perturbations, real state is confined to the ellipsoid. In particular, a minimization problem is given to determine the radius of the designed ellipsoid and the estimator gain matrix by testifying the feasibility of some recursive matrix inequalities. Finally, some simulations are used to show the feasibility of the developed non-fragile suboptimal state estimation strategy.

关键词： Convex optimization logarithmic quantization maximum-error-first protocol memristive neural networks mixed time-delays set-membership state estimation

来源：评论

学校读者我要写书评

暂无评论

Hybrid-driven H∞ filter design for T-S fuzzy systems with quantization

引用

NONLINEAR ANALYSIS-HYBRID SYSTEMS 2019年 31卷 135-152页

作者： Liu, Jinliang Wei, Lili Cao, Jie Fei, Shumin Nanjing Univ Finance & Econ Coll Informat Engn Nanjing 210023 Jiangsu Peoples R China Collaborat Innovat Ctr Modern Grain Circulat & Sa Nanjing 210023 Jiangsu Peoples R China Southeast Univ Sch Automat Nanjing 210096 Jiangsu Peoples R China

This paper is mainly concerned with hybrid-driven H-infinity filtering for a class of Takagi-Sugeno (T-S) fuzzy systems with quantization. To reduce the redundancy of transmission data and save the network bandwidth, a hybrid-driven scheme and a logarithmic quantizer are introduced in this paper. Firstly, by taking the effect of hybrid-driven scheme and quantization into consideration, a mathematical H-infinity filter model for T-S fuzzy systems is constructed. Secondly, by applying Lyapunov stability theory, sufficient conditions for asymptotical stabilization of desired system are obtained. Moreover, an explicit algorithm for H-infinity filter design is presented with the help of linear matrix inequality (LMI) techniques. Finally, numerical and physical simulations show the usefulness of the proposed filter design approach. (C) 2018 Elsevier Ltd. All rights reserved.

关键词： Filter design Hybrid-driven scheme T-S fuzzy systems logarithmic quantization

来源：评论

学校读者我要写书评

暂无评论

H_∞ state estimation for discrete memristive neural networks with signal quantization and probabilistic time delay

引用

SYSTEMS SCIENCE & CONTROL ENGINEERING 2021年第1期9卷 764-774页

作者： Feng, Le Zhao, Liang Ban, Liqun Harbin Univ Sci & Technol Dept Math Harbin Peoples R China Harbin Univ Sci & Technol Heilongjiang Prov Key Lab Optimizat Control & Int Harbin Peoples R China

In this paper, the problem of H-infinity state estimation is discussed for a class of delayed discrete memristive neural networks with signal quantization. A random variable obeying the Bernoulli distribution is used to describe the probabilistic time delay. A switching function is introduced to reflect the state dependence of memristive connection weight on neurons. Our aim is to design a state estimator to ensure that the specified disturbance attenuation level is guaranteed. By using Lyapunov stability theory and inequality scaling techniques, the specific explicit expression of gain parameter is given. Finally, a numerical example is given to verify the effectiveness of the proposed estimation method.

关键词： Memristive neural networks (MNNs) probabilistic time delay (PTD) logarithmic quantization H-infinity state estimation

来源：评论

学校读者我要写书评

暂无评论

Successive Log quantization for Cost-Efficient Neural Networks Using Stochastic Computing 19

Successive Log Quantization for Cost-Efficient Neural Networ...

引用

56th ACM/EDAC/IEEE Design Automation Conference (DAC)

作者： Lee, Sugil Sim, Hyeonuk Choi, Jooyeon Lee, Jongeun UNIST Sch Elect & Comp Engn Ulsan South Korea Seoul Natl Univ Neural Proc Res Ctr Seoul South Korea

ISBN: (纸本)9781450367257

Despite the multifaceted benefits of stochastic computing (SC) such as low cost, low power, and flexible precision, SC-based deep neural networks (DNNs) still suffer from the long-latency problem, especially for those with high precision requirements. While log quantization can be of help, it has its own accuracy-saturation problem due to uneven precision distribution. In this paper we propose successive log quantization (SLQ), which extends log quantization with significant improvements in precision and accuracy, and apply it to state-of-the-art SC-DNNs. SLQ reuses the existing datapath of log quantization, and thus retains its advantages such as simple multiplier hardware. Our experimental results demonstrate that our SLQ can significantly extend both the accuracy and efficiency of SC-DNNs over the state-of-the-art solutions, including linear-quantized and log-quantized SC-DNNs, achieving less than 1 similar to 1.5%p accuracy drop for AlexNet, SqueezeNet, and VGG-S at mere 4 similar to 5-bit weight resolution.

关键词： Deep neural network (DNN) logarithmic quantization Stochastic computing Adaptive precision

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：