To meet the requirements of quick positioning of mobile terminals from base stations(BSs)or third-party devices,as well as to improve the convergence speed and reduce the steady state maladjustment of the least mean s...To meet the requirements of quick positioning of mobile terminals from base stations(BSs)or third-party devices,as well as to improve the convergence speed and reduce the steady state maladjustment of the least mean square(LMS)method,a new logarithmic-sigmoid variable step-size LMS(LG-SVSLMS)was proposed and applied to estimate the direction of arrival(DOA)of orthogonal frequency division multiple access(OFDMA)signals.Based on the proposed LG-SVSLMS,a non-blind DOA estimation system for OFDMA signals was constructed.The proposed LG-SVSLMS adopts a new multi-parameter step-size update function which combines the sigmoid function and the logarithmic function.It controls the adjustment magnitude of step-size during the initial and steady state phases of the LMS method to achieve both a high convergence speed and low steady state maladjustment.Finally,simulation was conducted to verify the performance of the LG-SVSLMS.The simulation results show that the non-blind DOA estimation system based on the LG-SVSLMS can accurately estimate the DOA of the target signal in the scenario where interference signals from multi-source and multi-path fading signals arrive at the third-party devices asynchronously with the target signal,and the estimation deviation is within±3°.The non-blind DOA estimation for OFDMA signals with the proposed LG-SVSLMS is of great significance for the instant positioning technology of mobile terminals based on the adaptive antenna array.展开更多
To solve the contradiction between convergence rate and steady-state error in least mean square (LMS) algorithm, basing on independence assumption, this paper proposes and proves the optimal step-size theorem from the...To solve the contradiction between convergence rate and steady-state error in least mean square (LMS) algorithm, basing on independence assumption, this paper proposes and proves the optimal step-size theorem from the view of minimizing mean squared error (MSE). The theorem reveals the one-to-one mapping between the optimal step-size and MSE. Following the theorem, optimal variable step-size LMS (OVS-LMS) model, describing the theoretical bound of the convergence rate of LMS algorithm, is constructed. Then we discuss the selection of initial optimal step-size and updating of optimal step-size at the time of unknown system changing. At last an optimal step-size LMS algorithm is proposed and tested in various environments. Simulation results show the proposed algorithm is very close to the theoretical bound.展开更多
Implicit-explicit （IMEX） linear multistep methods are popular techniques for solving partial differential equations （PDEs） with terms of different types. While fixed timestep versions of such schemes have been dev...Implicit-explicit （IMEX） linear multistep methods are popular techniques for solving partial differential equations （PDEs） with terms of different types. While fixed timestep versions of such schemes have been developed and studied, implicit-explicit schemes also naturally arise in general situations where the temporal smoothness of the solution changes. In this paper we consider easily implementable variable step-size implicit-explicit （VSIMEX） linear multistep methods for time-dependent PDEs. Families of order-p, pstep VSIMEX schemes are constructed and analyzed, where p ranges from 1 to 4. The corresponding schemes are simple to implement and have the property that they reduce to the classical IMEX schemes whenever constant time step-sizes are imposed. The methods are validated on the Burgers＇ equation. These results demonstrate that by varying the time step-size, VSIMEX methods can outperform their fixed time step counterparts while still maintaining good numerical behavior.展开更多
In this work,we analyze the three-step backward differentiation formula(BDF3)method for solving the Allen-Cahn equation on variable grids.For BDF2 method,the discrete orthogonal convolution(DOC)kernels are positive,th...In this work,we analyze the three-step backward differentiation formula(BDF3)method for solving the Allen-Cahn equation on variable grids.For BDF2 method,the discrete orthogonal convolution(DOC)kernels are positive,the stability and convergence analysis are well established in[Liao and Zhang,Math.Comp.,90(2021),1207–1226]and[Chen,Yu,and Zhang,arXiv:2108.02910,2021].However,the numerical analysis for BDF3 method with variable steps seems to be highly nontrivial due to the additional degrees of freedom and the non-positivity of DOC kernels.By developing a novel spectral norm inequality,the unconditional stability and convergence are rigorously proved under the updated step ratio restriction rk:=τk/τk−1≤1.405 for BDF3 method.Finally,numerical experiments are performed to illustrate the theoretical results.To the best of our knowledge,this is the first theoretical analysis of variable steps BDF3 method for the Allen-Cahn equation.展开更多
With independence assumption, this paper proposes and proves the superior step-size theorem on least mean square (LMS) algorithm, from the view of minimizing mean squared error (MSE). Following the theorem we construc...With independence assumption, this paper proposes and proves the superior step-size theorem on least mean square (LMS) algorithm, from the view of minimizing mean squared error (MSE). Following the theorem we construct a parallel variable step-size LMS filters algorithm. The theoretical model of the proposed algorithm is analyzed in detail. Simulations show the proposed theoretical model is quite close to the optimal variable step-size LMS (OVS-LMS) model. The experimental learning curves of the proposed algorithm also show the fastest convergence and fine tracking performance. The proposed algorithm is therefore a good realization of the OVS-LMS model.展开更多
Acoustic echo cancellation is often applied in communication and video call system to reduce unnecessary echoes generated between speakers and microphones.In these systems,the speech input signal of the adaptive filte...Acoustic echo cancellation is often applied in communication and video call system to reduce unnecessary echoes generated between speakers and microphones.In these systems,the speech input signal of the adaptive filter is often colored and unstable,which decays the convergence rate of the adaptive filter if the NLMS algorithm is used.In this paper,an improved nonparametric variable step-size subband(NPVSS-NSAF)algorithm is proposed to address the problem.The variable step-size is derived by minimizing the sum of the square Euclidean norm of the difference between the optimal weight vectors to be updated and the past estimated weight vectors.Then the parameters are eliminated by using the power of subband signal noise equal to the power of subband posteriori error.The performance of the proposed algorithm is simulated in the aspects of misalignment and return loss enhancement.Experiment results show a fast convergence rate and low misalignment of the proposed algorithm in system identification.展开更多
This paper proposes a robust adaptive filter based on the exponent sin cost to improve the capability against Gaussian or multiple types of non-Gaussian noises of the adaptive filtering algorithm when dealing with tim...This paper proposes a robust adaptive filter based on the exponent sin cost to improve the capability against Gaussian or multiple types of non-Gaussian noises of the adaptive filtering algorithm when dealing with time-varying/time-invariant linear systems function exponent sin(ExpSin).Then a variable step-size(VSS)-ExpSin algorithm is extended further.Besides,the stepsize,the convergence,and the steady-state performance of the proposed algorithm are validated experimentally.The Monte Carlo simulation results of linear system identification illustrate the principle and efficiency of this proposed adaptive filtering algorithm.Results suggest that the proposed adaptive filtering algorithm has superior performance when estimating the unknown linear systems under multiple-types measurement noises.展开更多
An internal state variable(ISV)model was established according to the experimental results of hot plane strain compression(PSC)to predict the microstructure evolution during hot spinning of ZK61 alloy.The effects of t...An internal state variable(ISV)model was established according to the experimental results of hot plane strain compression(PSC)to predict the microstructure evolution during hot spinning of ZK61 alloy.The effects of the internal variables were considered in this ISV model,and the parameters were optimized by genetic algorithm.After validation,the ISV model was used to simulate the evolution of grain size(GS)and dynamic recrystallization(DRX)fraction during hot spinning via Abaqus and its subroutine Vumat.By comparing the simulated results with the experimental results,the application of the ISV model was proven to be reliable.Meanwhile,the strength of the thin-walled spun ZK61 tube increased from 303 to 334 MPa due to grain refinement by DRX and texture strengthening.Besides,some ultrafine grains(0.5μm)that played an important role in mechanical properties were formed due to the proliferation,movement,and entanglement of dislocations during the spinning process.展开更多
Large-scale Language Models(LLMs)have achieved significant breakthroughs in Natural Language Processing(NLP),driven by the pre-training and fine-tuning paradigm.While this approach allows models to specialize in speci...Large-scale Language Models(LLMs)have achieved significant breakthroughs in Natural Language Processing(NLP),driven by the pre-training and fine-tuning paradigm.While this approach allows models to specialize in specific tasks with reduced training costs,the substantial memory requirements during fine-tuning present a barrier to broader deployment.Parameter-Efficient Fine-Tuning(PEFT)techniques,such as Low-Rank Adaptation(LoRA),and parameter quantization methods have emerged as solutions to address these challenges by optimizing memory usage and computational efficiency.Among these,QLoRA,which combines PEFT and quantization,has demonstrated notable success in reducing memory footprints during fine-tuning,prompting the development of various QLoRA variants.Despite these advancements,the quantitative impact of key variables on the fine-tuning performance of quantized LLMs remains underexplored.This study presents a comprehensive analysis of these key variables,focusing on their influence across different layer types and depths within LLM architectures.Our investigation uncovers several critical findings:(1)Larger layers,such as MLP layers,can maintain performance despite reductions in adapter rank,while smaller layers,like self-attention layers,aremore sensitive to such changes;(2)The effectiveness of balancing factors depends more on specific values rather than layer type or depth;(3)In quantization-aware fine-tuning,larger layers can effectively utilize smaller adapters,whereas smaller layers struggle to do so.These insights suggest that layer type is a more significant determinant of fine-tuning success than layer depth when optimizing quantized LLMs.Moreover,for the same discount of trainable parameters,reducing the trainable parameters in a larger layer is more effective in preserving fine-tuning accuracy than in a smaller one.This study provides valuable guidance for more efficient fine-tuning strategies and opens avenues for further research into optimizing LLM fine-tuning in resource-constrained environments.展开更多
We consider a single server constant retrial queue,in which a state-dependent service policy is used to control the service rate.Customer arrival follows Poisson process,while service time and retrial time are exponen...We consider a single server constant retrial queue,in which a state-dependent service policy is used to control the service rate.Customer arrival follows Poisson process,while service time and retrial time are exponential distributions.Whenever the server is available,it admits the retrial customers into service based on a first-come first-served rule.The service rate adjusts in real-time based on the retrial queue length.An iterative algorithm is proposed to numerically solve the personal optimal problem in the fully observable scenario.Furthermore,we investigate the impact of parameters on the social optimal threshold.The effectiveness of the results is illustrated by two examples.展开更多
Fiber-reinforced composites are an ideal material for the lightweight design of aerospace structures. Especially in recent years, with the rapid development of composite additive manufacturing technology, the design o...Fiber-reinforced composites are an ideal material for the lightweight design of aerospace structures. Especially in recent years, with the rapid development of composite additive manufacturing technology, the design optimization of variable stiffness of fiber-reinforced composite laminates has attracted widespread attention from scholars and industry. In these aerospace composite structures, numerous cutout panels and shells serve as access points for maintaining electrical, fuel, and hydraulic systems. The traditional fiber-reinforced composite laminate subtractive drilling manufacturing inevitably faces the problems of interlayer delamination, fiber fracture, and burr of the laminate. Continuous fiber additive manufacturing technology offers the potential for integrated design optimization and manufacturing with high structural performance. Considering the integration of design and manufacturability in continuous fiber additive manufacturing, the paper proposes linear and nonlinear filtering strategies based on the Normal Distribution Fiber Optimization (NDFO) material interpolation scheme to overcome the challenge of discrete fiber optimization results, which are difficult to apply directly to continuous fiber additive manufacturing. With minimizing structural compliance as the objective function, the proposed approach provides a strategy to achieve continuity of discrete fiber paths in the variable stiffness design optimization of composite laminates with regular and irregular holes. In the variable stiffness design optimization model, the number of candidate fiber laying angles in the NDFO material interpolation scheme is considered as design variable. The sensitivity information of structural compliance with respect to the number of candidate fiber laying angles is obtained using the analytical sensitivity analysis method. Based on the proposed variable stiffness design optimization method for complex perforated composite laminates, the numerical examples consider the variable stiffness design optimization of typical non-perforated and perforated composite laminates with circular, square, and irregular holes, and systematically discuss the number of candidate discrete fiber laying angles, discrete fiber continuous filtering strategies, and filter radius on structural compliance, continuity, and manufacturability. The optimized discrete fiber angles of variable stiffness laminates are converted into continuous fiber laying paths using a streamlined process for continuous fiber additive manufacturing. Meanwhile, the optimized non-perforated and perforated MBB beams after discrete fiber continuous treatment, are manufactured using continuous fiber co-extrusion additive manufacturing technology to verify the effectiveness of the variable stiffness fiber optimization framework proposed in this paper.展开更多
The complete convergence for weighted sums of sequences of independent,identically distributed random variables under sublinear expectation space is studied.By moment inequality and truncation methods,we establish the...The complete convergence for weighted sums of sequences of independent,identically distributed random variables under sublinear expectation space is studied.By moment inequality and truncation methods,we establish the equivalent conditions of complete convergence for weighted sums of sequences of independent,identically distributed random variables under sublinear expectation space.The results complement the corresponding results in probability space to those for sequences of independent,identically distributed random variables under sublinear expectation space.展开更多
With the application of 2.5D Woven Variable Thickness Composites(2.5DWVTC)in aviation and other fields,the issue of strength failure in this composite type has become a focal point.First,a three-step modeling approach...With the application of 2.5D Woven Variable Thickness Composites(2.5DWVTC)in aviation and other fields,the issue of strength failure in this composite type has become a focal point.First,a three-step modeling approach is proposed to rapidly construct full-scale meso-finite element models for Outer Reduction Yarn Woven Composites(ORYWC)and Inner Reduction Yarn Woven Composites(IRYWC).Then,six independent damage variables are identified:yarn fiber tension/compression,yarn matrix tension/compression,and resin matrix tension/compression.These variables are utilized to establish the constitutive equation of woven composites,considering the coupling effects of microscopic damage.Finally,combined with the Hashin failure criterion and von Mises failure criterion,the strength prediction model is implemented in ANSYS using APDL language to simulate the strength failure process of 2.5DWVTC.The results show that the predicted stiffness and strength values of various parts of ORYWC and IRYWC are in good agreement with the relevant test results.展开更多
In this paper,by utilizing the Marcinkiewicz-Zygmund inequality and Rosenthal-type inequality of negatively superadditive dependent(NSD)random arrays and truncated method,we investigate the complete f-moment convergen...In this paper,by utilizing the Marcinkiewicz-Zygmund inequality and Rosenthal-type inequality of negatively superadditive dependent(NSD)random arrays and truncated method,we investigate the complete f-moment convergence of NSD random variables.We establish and improve a general result on the complete f-moment convergence for Sung’s type randomly weighted sums of NSD random variables under some general assumptions.As an application,we show the complete consistency for the randomly weighted estimator in a nonparametric regression model based on NSD errors.展开更多
We investigate a class of elliptic equations with an L^(1)source in the framework of variable exponent spaces.A key characteristic of these equations is the coexistence of a degenerate coercivity term and a lower-orde...We investigate a class of elliptic equations with an L^(1)source in the framework of variable exponent spaces.A key characteristic of these equations is the coexistence of a degenerate coercivity term and a lower-order convection term.By employing innovative integralbased test functions,we derive the necessary a priori estimates.To prove the convergence of solutions to the degenerate coercivity problem,we adopt a method that combines monotonicity and truncation techniques.This approach allows us to demonstrate that the gradient sequences converge almost everywhere.展开更多
In this paper we present certain bilinear estimates for commutators on Besov spaces with variable smoothness and integrability,and under no vanishing assumptions on the divergence of vector fields.Such commutator esti...In this paper we present certain bilinear estimates for commutators on Besov spaces with variable smoothness and integrability,and under no vanishing assumptions on the divergence of vector fields.Such commutator estimates are motivated by the study of well-posedness results for some models in incompressible fuid mechanics.展开更多
The development of chassis active safety control technology has improved vehicle stability under extreme conditions.However,its cross-system and multi-functional characteristics make the controller difficult to achiev...The development of chassis active safety control technology has improved vehicle stability under extreme conditions.However,its cross-system and multi-functional characteristics make the controller difficult to achieve cooperative goals.In addition,the chassis system,which has high complexity,numerous subsystems,and strong coupling,will also lead to low computing efficiency and poor control effect of the controller.Therefore,this paper proposes a scenario-driven hybrid distributed model predictive control algorithm with variable control topology.This algorithm divides multiple stability regions based on the vehicle’s β−γ phase plane,forming a mapping relationship between the control structure and the vehicle’s state.A control input fusion mechanism within the transition domain is designed to mitigate the problems of system state oscillation and control input jitter caused by switching control structures.Then,a distributed state-space equation with state coupling and input coupling characteristics is constructed,and a weighted local agent cost function in quadratic programming is derived.Through cost coupling,local agents can coordinate global performance goals.Finally,through Simulink/CarSim joint simulation and hardware-in-the-loop(HIL)test,the proposed algorithm is validated to improve vehicle stability while ensuring trajectory tracking accuracy and has good applicability for multi-objective coordinated control.This paper combines the advantages of distributed MPC and decentralized MPC,achieving a balance between approximating the global optimal results and the solution’s efficiency.展开更多
Cold seeps are oases for biological communities on the sea floor around hydrocarbon emission pathways.Microbial utilization of methane and other hydrocarbons yield products that fuel rich chemosynthetic communities at...Cold seeps are oases for biological communities on the sea floor around hydrocarbon emission pathways.Microbial utilization of methane and other hydrocarbons yield products that fuel rich chemosynthetic communities at these sites.One such site in the cold seep ecosystem of Krishna-Godavari basin(K-G basin)along the east coast of India,discovered in Feb 2018 at a depth of 1800 m was assessed for its bacterial diversity.The seep bacterial communities were dominated by phylum Proteobacteria(57%),Firmicutes(16%)and unclassified species belonging to the family Helicobacteriaceae.The surface sediments of the seep had maximum OTUs(operational taxonomic units)(2.27×10^(3))with a Shannon alpha diversity index of 8.06.In general,environmental parameters like total organic carbon(p<0.01),sulfate(p<0.001),sulfide(p<0.05)and methane(p<0.01)were responsible for shaping the bacterial community of the cold seep ecosystem in the K-G Basin.Environmental parameters play a significant role in changing the bacterial diversity richness between different cold seep environments in the oceans.展开更多
This paper presents a continuum manipulator inspired by the anatomical characteristics of the elephant trunk.Specifically,the manipulator mimics the conoid profile of the elephant trunk,which helps to enhance its stre...This paper presents a continuum manipulator inspired by the anatomical characteristics of the elephant trunk.Specifically,the manipulator mimics the conoid profile of the elephant trunk,which helps to enhance its strength.The design features two concentric parts:inner pneumatically actuated bellows and an outer tendon-driven helical spring.The tendons control the omnidirectional bending of the manipulator,while the fusion of the pneumatic bellows with the tendon-driven spring results in an antagonistic actuation mechanism that provides the manipulator with variable stiffness and extensibility.This paper presents a new design for extensible manipulator and analyzes its stiffness and motion characteristics.Experimental results are consistent with theoretical analysis,thereby demonstrating the validity of the theoretical approach and the versatile practical mechanical properties of the continuum manipulator.The impressive extensibility and variable stiffness of the manipulator were further demonstrated by performing a pin-hole assembly task.展开更多
Variable selection identifies the best subset of covariates when building the prediction model,among all possible subsets.In this paper,we propose a novel reinforced variable selection method,called‘Actor-Critic-Pred...Variable selection identifies the best subset of covariates when building the prediction model,among all possible subsets.In this paper,we propose a novel reinforced variable selection method,called‘Actor-Critic-Predictor’.The actor takes an action to choose variables and the predictor evaluates the action based on a well-designed reward function,where the critic learns the reward baseline.We model the variable selection process as a multi-armed bandit and update the subset of selected variables using a natural policy gradient algorithm.We provide an analytical framework on how different errors impact the performance of our method theoretically.Large amounts of experiments on synthetic and real datasets show that the proposed framework is easily implemented and outperforms classical variable selection methods in a wide range of scenarios.展开更多
基金The Social Development Projects of Jiangsu Science and Technology Department(No.BE2018704)the Technological Innovation Projects of Ministry of Public Security of China(No.20170001)。
文摘To meet the requirements of quick positioning of mobile terminals from base stations(BSs)or third-party devices,as well as to improve the convergence speed and reduce the steady state maladjustment of the least mean square(LMS)method,a new logarithmic-sigmoid variable step-size LMS(LG-SVSLMS)was proposed and applied to estimate the direction of arrival(DOA)of orthogonal frequency division multiple access(OFDMA)signals.Based on the proposed LG-SVSLMS,a non-blind DOA estimation system for OFDMA signals was constructed.The proposed LG-SVSLMS adopts a new multi-parameter step-size update function which combines the sigmoid function and the logarithmic function.It controls the adjustment magnitude of step-size during the initial and steady state phases of the LMS method to achieve both a high convergence speed and low steady state maladjustment.Finally,simulation was conducted to verify the performance of the LG-SVSLMS.The simulation results show that the non-blind DOA estimation system based on the LG-SVSLMS can accurately estimate the DOA of the target signal in the scenario where interference signals from multi-source and multi-path fading signals arrive at the third-party devices asynchronously with the target signal,and the estimation deviation is within±3°.The non-blind DOA estimation for OFDMA signals with the proposed LG-SVSLMS is of great significance for the instant positioning technology of mobile terminals based on the adaptive antenna array.
基金This work was supported in part by the National Fundamental Research Program(Grant No.G1998030406)the National Natural Science Foundation of China(Grant No.69972020)by the State Key Lab on Microwave and Digital Communications,Department of Electronics Engineering,Tsinghua University.
文摘To solve the contradiction between convergence rate and steady-state error in least mean square (LMS) algorithm, basing on independence assumption, this paper proposes and proves the optimal step-size theorem from the view of minimizing mean squared error (MSE). The theorem reveals the one-to-one mapping between the optimal step-size and MSE. Following the theorem, optimal variable step-size LMS (OVS-LMS) model, describing the theoretical bound of the convergence rate of LMS algorithm, is constructed. Then we discuss the selection of initial optimal step-size and updating of optimal step-size at the time of unknown system changing. At last an optimal step-size LMS algorithm is proposed and tested in various environments. Simulation results show the proposed algorithm is very close to the theoretical bound.
基金supported by an NSERC Canada Postgraduate Scholarshipsupported by a grant from NSERC Canada
文摘Implicit-explicit （IMEX） linear multistep methods are popular techniques for solving partial differential equations （PDEs） with terms of different types. While fixed timestep versions of such schemes have been developed and studied, implicit-explicit schemes also naturally arise in general situations where the temporal smoothness of the solution changes. In this paper we consider easily implementable variable step-size implicit-explicit （VSIMEX） linear multistep methods for time-dependent PDEs. Families of order-p, pstep VSIMEX schemes are constructed and analyzed, where p ranges from 1 to 4. The corresponding schemes are simple to implement and have the property that they reduce to the classical IMEX schemes whenever constant time step-sizes are imposed. The methods are validated on the Burgers＇ equation. These results demonstrate that by varying the time step-size, VSIMEX methods can outperform their fixed time step counterparts while still maintaining good numerical behavior.
基金supported by the Science Fund for Distinguished Young Scholars of Gansu Province(Grant No.23JRRA1020)the Fundamental Research Funds for the Central Universities(Grant No.lzujbky-2023-06).
文摘In this work,we analyze the three-step backward differentiation formula(BDF3)method for solving the Allen-Cahn equation on variable grids.For BDF2 method,the discrete orthogonal convolution(DOC)kernels are positive,the stability and convergence analysis are well established in[Liao and Zhang,Math.Comp.,90(2021),1207–1226]and[Chen,Yu,and Zhang,arXiv:2108.02910,2021].However,the numerical analysis for BDF3 method with variable steps seems to be highly nontrivial due to the additional degrees of freedom and the non-positivity of DOC kernels.By developing a novel spectral norm inequality,the unconditional stability and convergence are rigorously proved under the updated step ratio restriction rk:=τk/τk−1≤1.405 for BDF3 method.Finally,numerical experiments are performed to illustrate the theoretical results.To the best of our knowledge,this is the first theoretical analysis of variable steps BDF3 method for the Allen-Cahn equation.
文摘With independence assumption, this paper proposes and proves the superior step-size theorem on least mean square (LMS) algorithm, from the view of minimizing mean squared error (MSE). Following the theorem we construct a parallel variable step-size LMS filters algorithm. The theoretical model of the proposed algorithm is analyzed in detail. Simulations show the proposed theoretical model is quite close to the optimal variable step-size LMS (OVS-LMS) model. The experimental learning curves of the proposed algorithm also show the fastest convergence and fine tracking performance. The proposed algorithm is therefore a good realization of the OVS-LMS model.
基金This work was supported by the National Key Research and Development Program of China(Grant No.2018YFF0213602).
文摘Acoustic echo cancellation is often applied in communication and video call system to reduce unnecessary echoes generated between speakers and microphones.In these systems,the speech input signal of the adaptive filter is often colored and unstable,which decays the convergence rate of the adaptive filter if the NLMS algorithm is used.In this paper,an improved nonparametric variable step-size subband(NPVSS-NSAF)algorithm is proposed to address the problem.The variable step-size is derived by minimizing the sum of the square Euclidean norm of the difference between the optimal weight vectors to be updated and the past estimated weight vectors.Then the parameters are eliminated by using the power of subband signal noise equal to the power of subband posteriori error.The performance of the proposed algorithm is simulated in the aspects of misalignment and return loss enhancement.Experiment results show a fast convergence rate and low misalignment of the proposed algorithm in system identification.
文摘This paper proposes a robust adaptive filter based on the exponent sin cost to improve the capability against Gaussian or multiple types of non-Gaussian noises of the adaptive filtering algorithm when dealing with time-varying/time-invariant linear systems function exponent sin(ExpSin).Then a variable step-size(VSS)-ExpSin algorithm is extended further.Besides,the stepsize,the convergence,and the steady-state performance of the proposed algorithm are validated experimentally.The Monte Carlo simulation results of linear system identification illustrate the principle and efficiency of this proposed adaptive filtering algorithm.Results suggest that the proposed adaptive filtering algorithm has superior performance when estimating the unknown linear systems under multiple-types measurement noises.
基金supported by the National Natural Science Foundation of China(No.51905123)Major Scientific and Technological Innovation Program of Shandong Province,China(Nos.2020CXGC010303,2022ZLGX04)Key R&D Programme of Shandong Province,China(No.2022JMRH0308).
文摘An internal state variable(ISV)model was established according to the experimental results of hot plane strain compression(PSC)to predict the microstructure evolution during hot spinning of ZK61 alloy.The effects of the internal variables were considered in this ISV model,and the parameters were optimized by genetic algorithm.After validation,the ISV model was used to simulate the evolution of grain size(GS)and dynamic recrystallization(DRX)fraction during hot spinning via Abaqus and its subroutine Vumat.By comparing the simulated results with the experimental results,the application of the ISV model was proven to be reliable.Meanwhile,the strength of the thin-walled spun ZK61 tube increased from 303 to 334 MPa due to grain refinement by DRX and texture strengthening.Besides,some ultrafine grains(0.5μm)that played an important role in mechanical properties were formed due to the proliferation,movement,and entanglement of dislocations during the spinning process.
基金supported by the National Key R&D Program of China(No.2021YFB0301200)National Natural Science Foundation of China(No.62025208).
文摘Large-scale Language Models(LLMs)have achieved significant breakthroughs in Natural Language Processing(NLP),driven by the pre-training and fine-tuning paradigm.While this approach allows models to specialize in specific tasks with reduced training costs,the substantial memory requirements during fine-tuning present a barrier to broader deployment.Parameter-Efficient Fine-Tuning(PEFT)techniques,such as Low-Rank Adaptation(LoRA),and parameter quantization methods have emerged as solutions to address these challenges by optimizing memory usage and computational efficiency.Among these,QLoRA,which combines PEFT and quantization,has demonstrated notable success in reducing memory footprints during fine-tuning,prompting the development of various QLoRA variants.Despite these advancements,the quantitative impact of key variables on the fine-tuning performance of quantized LLMs remains underexplored.This study presents a comprehensive analysis of these key variables,focusing on their influence across different layer types and depths within LLM architectures.Our investigation uncovers several critical findings:(1)Larger layers,such as MLP layers,can maintain performance despite reductions in adapter rank,while smaller layers,like self-attention layers,aremore sensitive to such changes;(2)The effectiveness of balancing factors depends more on specific values rather than layer type or depth;(3)In quantization-aware fine-tuning,larger layers can effectively utilize smaller adapters,whereas smaller layers struggle to do so.These insights suggest that layer type is a more significant determinant of fine-tuning success than layer depth when optimizing quantized LLMs.Moreover,for the same discount of trainable parameters,reducing the trainable parameters in a larger layer is more effective in preserving fine-tuning accuracy than in a smaller one.This study provides valuable guidance for more efficient fine-tuning strategies and opens avenues for further research into optimizing LLM fine-tuning in resource-constrained environments.
基金supported by the National Natural Science Foundation of China(Grant No.11971486)。
文摘We consider a single server constant retrial queue,in which a state-dependent service policy is used to control the service rate.Customer arrival follows Poisson process,while service time and retrial time are exponential distributions.Whenever the server is available,it admits the retrial customers into service based on a first-come first-served rule.The service rate adjusts in real-time based on the retrial queue length.An iterative algorithm is proposed to numerically solve the personal optimal problem in the fully observable scenario.Furthermore,we investigate the impact of parameters on the social optimal threshold.The effectiveness of the results is illustrated by two examples.
基金supports for this research were provided by the National Natural Science Foundation of China(No.12272301,12002278,U1906233)the Guangdong Basic and Applied Basic Research Foundation,China(Nos.2023A1515011970,2024A1515010256)+1 种基金the Dalian City Supports Innovation and Entrepreneurship Projects for High-Level Talents,China(2021RD16)the Key R&D Project of CSCEC,China(No.CSCEC-2020-Z-4).
文摘Fiber-reinforced composites are an ideal material for the lightweight design of aerospace structures. Especially in recent years, with the rapid development of composite additive manufacturing technology, the design optimization of variable stiffness of fiber-reinforced composite laminates has attracted widespread attention from scholars and industry. In these aerospace composite structures, numerous cutout panels and shells serve as access points for maintaining electrical, fuel, and hydraulic systems. The traditional fiber-reinforced composite laminate subtractive drilling manufacturing inevitably faces the problems of interlayer delamination, fiber fracture, and burr of the laminate. Continuous fiber additive manufacturing technology offers the potential for integrated design optimization and manufacturing with high structural performance. Considering the integration of design and manufacturability in continuous fiber additive manufacturing, the paper proposes linear and nonlinear filtering strategies based on the Normal Distribution Fiber Optimization (NDFO) material interpolation scheme to overcome the challenge of discrete fiber optimization results, which are difficult to apply directly to continuous fiber additive manufacturing. With minimizing structural compliance as the objective function, the proposed approach provides a strategy to achieve continuity of discrete fiber paths in the variable stiffness design optimization of composite laminates with regular and irregular holes. In the variable stiffness design optimization model, the number of candidate fiber laying angles in the NDFO material interpolation scheme is considered as design variable. The sensitivity information of structural compliance with respect to the number of candidate fiber laying angles is obtained using the analytical sensitivity analysis method. Based on the proposed variable stiffness design optimization method for complex perforated composite laminates, the numerical examples consider the variable stiffness design optimization of typical non-perforated and perforated composite laminates with circular, square, and irregular holes, and systematically discuss the number of candidate discrete fiber laying angles, discrete fiber continuous filtering strategies, and filter radius on structural compliance, continuity, and manufacturability. The optimized discrete fiber angles of variable stiffness laminates are converted into continuous fiber laying paths using a streamlined process for continuous fiber additive manufacturing. Meanwhile, the optimized non-perforated and perforated MBB beams after discrete fiber continuous treatment, are manufactured using continuous fiber co-extrusion additive manufacturing technology to verify the effectiveness of the variable stiffness fiber optimization framework proposed in this paper.
基金supported by Doctoral Scientific Research Starting Foundation of Jingdezhen Ceramic University(Grant No.102/01003002031)Re-accompanying Funding Project of Academic Achievements of Jingdezhen Ceramic University(Grant Nos.215/20506277,215/20506341)。
文摘The complete convergence for weighted sums of sequences of independent,identically distributed random variables under sublinear expectation space is studied.By moment inequality and truncation methods,we establish the equivalent conditions of complete convergence for weighted sums of sequences of independent,identically distributed random variables under sublinear expectation space.The results complement the corresponding results in probability space to those for sequences of independent,identically distributed random variables under sublinear expectation space.
基金supported by National Science and Technology Major Project,China(No.2017-IV-0007-0044)National Natural Science Foundation of China(No.52175142),National Natural Science Foundation of China(No.52305170)Natural Science Foundation of Sichuan Province,China(No.2022NSFSC1885)。
文摘With the application of 2.5D Woven Variable Thickness Composites(2.5DWVTC)in aviation and other fields,the issue of strength failure in this composite type has become a focal point.First,a three-step modeling approach is proposed to rapidly construct full-scale meso-finite element models for Outer Reduction Yarn Woven Composites(ORYWC)and Inner Reduction Yarn Woven Composites(IRYWC).Then,six independent damage variables are identified:yarn fiber tension/compression,yarn matrix tension/compression,and resin matrix tension/compression.These variables are utilized to establish the constitutive equation of woven composites,considering the coupling effects of microscopic damage.Finally,combined with the Hashin failure criterion and von Mises failure criterion,the strength prediction model is implemented in ANSYS using APDL language to simulate the strength failure process of 2.5DWVTC.The results show that the predicted stiffness and strength values of various parts of ORYWC and IRYWC are in good agreement with the relevant test results.
基金supported by the National Social Science Fundation(Grant No.21BTJ040)the Project of Outstanding Young People in University of Anhui Province(Grant Nos.2023AH020037,SLXY2024A001).
文摘In this paper,by utilizing the Marcinkiewicz-Zygmund inequality and Rosenthal-type inequality of negatively superadditive dependent(NSD)random arrays and truncated method,we investigate the complete f-moment convergence of NSD random variables.We establish and improve a general result on the complete f-moment convergence for Sung’s type randomly weighted sums of NSD random variables under some general assumptions.As an application,we show the complete consistency for the randomly weighted estimator in a nonparametric regression model based on NSD errors.
基金Supported by the National Natural Science Foundation of China(Grant No.11901131)。
文摘We investigate a class of elliptic equations with an L^(1)source in the framework of variable exponent spaces.A key characteristic of these equations is the coexistence of a degenerate coercivity term and a lower-order convection term.By employing innovative integralbased test functions,we derive the necessary a priori estimates.To prove the convergence of solutions to the degenerate coercivity problem,we adopt a method that combines monotonicity and truncation techniques.This approach allows us to demonstrate that the gradient sequences converge almost everywhere.
文摘In this paper we present certain bilinear estimates for commutators on Besov spaces with variable smoothness and integrability,and under no vanishing assumptions on the divergence of vector fields.Such commutator estimates are motivated by the study of well-posedness results for some models in incompressible fuid mechanics.
基金Supported by National Natural Science Foundation of China(Grant Nos.52225212,52272418,U22A20100)National Key Research and Development Program of China(Grant No.2022YFB2503302).
文摘The development of chassis active safety control technology has improved vehicle stability under extreme conditions.However,its cross-system and multi-functional characteristics make the controller difficult to achieve cooperative goals.In addition,the chassis system,which has high complexity,numerous subsystems,and strong coupling,will also lead to low computing efficiency and poor control effect of the controller.Therefore,this paper proposes a scenario-driven hybrid distributed model predictive control algorithm with variable control topology.This algorithm divides multiple stability regions based on the vehicle’s β−γ phase plane,forming a mapping relationship between the control structure and the vehicle’s state.A control input fusion mechanism within the transition domain is designed to mitigate the problems of system state oscillation and control input jitter caused by switching control structures.Then,a distributed state-space equation with state coupling and input coupling characteristics is constructed,and a weighted local agent cost function in quadratic programming is derived.Through cost coupling,local agents can coordinate global performance goals.Finally,through Simulink/CarSim joint simulation and hardware-in-the-loop(HIL)test,the proposed algorithm is validated to improve vehicle stability while ensuring trajectory tracking accuracy and has good applicability for multi-objective coordinated control.This paper combines the advantages of distributed MPC and decentralized MPC,achieving a balance between approximating the global optimal results and the solution’s efficiency.
文摘Cold seeps are oases for biological communities on the sea floor around hydrocarbon emission pathways.Microbial utilization of methane and other hydrocarbons yield products that fuel rich chemosynthetic communities at these sites.One such site in the cold seep ecosystem of Krishna-Godavari basin(K-G basin)along the east coast of India,discovered in Feb 2018 at a depth of 1800 m was assessed for its bacterial diversity.The seep bacterial communities were dominated by phylum Proteobacteria(57%),Firmicutes(16%)and unclassified species belonging to the family Helicobacteriaceae.The surface sediments of the seep had maximum OTUs(operational taxonomic units)(2.27×10^(3))with a Shannon alpha diversity index of 8.06.In general,environmental parameters like total organic carbon(p<0.01),sulfate(p<0.001),sulfide(p<0.05)and methane(p<0.01)were responsible for shaping the bacterial community of the cold seep ecosystem in the K-G Basin.Environmental parameters play a significant role in changing the bacterial diversity richness between different cold seep environments in the oceans.
基金supported by the National Key R&D Program of China(No.2018YFB1305400)the Major Research Plan of the National Natural Science Foundation of China(No.92048301)+1 种基金the National Natural Science Foundation of China(No.52025054)the Joint Research Fund between the National Natural Science Foundation of China(NSFC)and Shen Zhen(No.U1713201).
文摘This paper presents a continuum manipulator inspired by the anatomical characteristics of the elephant trunk.Specifically,the manipulator mimics the conoid profile of the elephant trunk,which helps to enhance its strength.The design features two concentric parts:inner pneumatically actuated bellows and an outer tendon-driven helical spring.The tendons control the omnidirectional bending of the manipulator,while the fusion of the pneumatic bellows with the tendon-driven spring results in an antagonistic actuation mechanism that provides the manipulator with variable stiffness and extensibility.This paper presents a new design for extensible manipulator and analyzes its stiffness and motion characteristics.Experimental results are consistent with theoretical analysis,thereby demonstrating the validity of the theoretical approach and the versatile practical mechanical properties of the continuum manipulator.The impressive extensibility and variable stiffness of the manipulator were further demonstrated by performing a pin-hole assembly task.
基金supported by the National Natural Science Foundation of China[grant numbers 12401387,11771268,12001356]the Fujian Alliance of Mathematics[grand number 2024SXLMQN01]the Talent Fund Project of Fuzhou University[grand number XRC-24048].
文摘Variable selection identifies the best subset of covariates when building the prediction model,among all possible subsets.In this paper,we propose a novel reinforced variable selection method,called‘Actor-Critic-Predictor’.The actor takes an action to choose variables and the predictor evaluates the action based on a well-designed reward function,where the critic learns the reward baseline.We model the variable selection process as a multi-armed bandit and update the subset of selected variables using a natural policy gradient algorithm.We provide an analytical framework on how different errors impact the performance of our method theoretically.Large amounts of experiments on synthetic and real datasets show that the proposed framework is easily implemented and outperforms classical variable selection methods in a wide range of scenarios.
