This paper introduces a computational cognitive architecture that serves as a comprehensive computational theory of the human mind,from cognitive science and computational psychology.The cognitive architecture(named C...This paper introduces a computational cognitive architecture that serves as a comprehensive computational theory of the human mind,from cognitive science and computational psychology.The cognitive architecture(named Clarion)has been justified by,and validated against,psychological data,findings,and theoretical constructs.One important theoretical background for it is the dual-process theories,which led to its overall two-level structuring in a hybrid neuro-symbolic way.Furthermore,given the recent advances in AI and computing technology,LLMs are being incorporated into the model to better capture human intuition and instinct(and implicit processes in general),in order to further enhance Clarion.Integrating Clarion and LLMs can also help to develop AI systems that are more capable,more reliable,and more human-like.Overall,the paper advocates a multidisciplinary approach towards developing better models for cognitive science and for AI.展开更多
A complete examination of Large Language Models’strengths,problems,and applications is needed due to their rising use across disciplines.Current studies frequently focus on single-use situations and lack a comprehens...A complete examination of Large Language Models’strengths,problems,and applications is needed due to their rising use across disciplines.Current studies frequently focus on single-use situations and lack a comprehensive understanding of LLM architectural performance,strengths,and weaknesses.This gap precludes finding the appropriate models for task-specific applications and limits awareness of emerging LLM optimization and deployment strategies.In this research,50 studies on 25+LLMs,including GPT-3,GPT-4,Claude 3.5,DeepKet,and hybrid multimodal frameworks like ContextDET and GeoRSCLIP,are thoroughly reviewed.We propose LLM application taxonomy by grouping techniques by task focus—healthcare,chemistry,sentiment analysis,agent-based simulations,and multimodal integration.Advanced methods like parameter-efficient tuning(LoRA),quantumenhanced embeddings(DeepKet),retrieval-augmented generation(RAG),and safety-focused models(GalaxyGPT)are evaluated for dataset requirements,computational efficiency,and performance measures.Frameworks for ethical issues,data limited hallucinations,and KDGI-enhanced fine-tuning like Woodpecker’s post-remedy corrections are highlighted.The investigation’s scope,mad,and methods are described,but the primary results are not.The work reveals that domain-specialized fine-tuned LLMs employing RAG and quantum-enhanced embeddings performbetter for context-heavy applications.In medical text normalization,ChatGPT-4 outperforms previous models,while two multimodal frameworks,GeoRSCLIP,increase remote sensing.Parameter-efficient tuning technologies like LoRA have minimal computing cost and similar performance,demonstrating the necessity for adaptive models in multiple domains.To discover the optimum domain-specific models,explain domain-specific fine-tuning,and present quantum andmultimodal LLMs to address scalability and cross-domain issues.The framework helps academics and practitioners identify,adapt,and innovate LLMs for different purposes.This work advances the field of efficient,interpretable,and ethical LLM application research.展开更多
The widespread proliferation of modern wireless devices coupled with overlapping power emissions has brought about electromagnetic(EM)pollution issues,posing many challenges to environment and human health.Therefore,t...The widespread proliferation of modern wireless devices coupled with overlapping power emissions has brought about electromagnetic(EM)pollution issues,posing many challenges to environment and human health.Therefore,the development of EM shielding devices with high green shielding index(gs)is essential,as they offer absorption-dominant protection that minimizes reflections and safeguards both health and electronics.MXene,with its intrinsic ultra-high electrical conductivity,liquid-phase tunable surface chemistry,low density,large specific surface area,thermal stability,and mechanical stability,has become the leading two-dimensional(2D)material driving the development of green EM shielding devices.In this review we emphasize device-level strategies with engineered architectures for MXene-based green EM shielding.We first examine MXene’s crystal and electronic structure and the fundamental attenuation mechanisms in MXene-based devices.Then we survey fabrication and assembly methods,analyzing three device-level strategies for MXene-based green EM shielded devices:3D architectures,metastructure/meta-surfaces,and external stimulus.Throughout,we highlight how MXene’s distinguished properties enable green EM interference(EMI)shielding devices that minimize secondary interference.Finally,we discuss the challenges faced in the effective utilization of MXene-based in green EM shielding devices,provide insights into these challenges,and offer guidelines for developing the solutions of next-generation green MXene-based EM shielding devices.展开更多
The network on chip（NoC）is used as a solution for the communication problems in a complex system on chip（SoC）design.To further enhance performances,the NoC architectures,a high level modeling and an evaluation met...The network on chip（NoC）is used as a solution for the communication problems in a complex system on chip（SoC）design.To further enhance performances,the NoC architectures,a high level modeling and an evaluation method based on OPNET are proposed to analyze their performances on different injection rates and traffic patterns.Simulation results for general NoC in terms of the average latency and the throughput are analyzed and used as a guideline to make appropriate choices for a given application.Finally,a MPEG4 decoder is mapped on different NoC architectures.Results prove the effectiveness of the evaluation method.展开更多
In this paper we discuss policy iteration methods for approximate solution of a finite-state discounted Markov decision problem, with a focus on feature-based aggregation methods and their connection with deep reinfor...In this paper we discuss policy iteration methods for approximate solution of a finite-state discounted Markov decision problem, with a focus on feature-based aggregation methods and their connection with deep reinforcement learning schemes. We introduce features of the states of the original problem, and we formulate a smaller "aggregate" Markov decision problem, whose states relate to the features. We discuss properties and possible implementations of this type of aggregation, including a new approach to approximate policy iteration. In this approach the policy improvement operation combines feature-based aggregation with feature construction using deep neural networks or other calculations. We argue that the cost function of a policy may be approximated much more accurately by the nonlinear function of the features provided by aggregation, than by the linear function of the features provided by neural networkbased reinforcement learning, thereby potentially leading to more effective policy improvement.展开更多
The rapid consumption of fossil fuel and increased environmental damage caused by it have given a strong impetus to the growth and development of fuelefficient vehicles. Hybrid electric vehicles (HEVs) have evolved fr...The rapid consumption of fossil fuel and increased environmental damage caused by it have given a strong impetus to the growth and development of fuelefficient vehicles. Hybrid electric vehicles (HEVs) have evolved from their inchoate state and are proving to be a promising solution to the serious existential problem posed to the planet earth. Not only do HEVs provide better fuel economy and lower emissions satisfying environmental legislations, but also they dampen the effect of rising fuel prices on consumers. HEVs combine the drive powers of an internal combustion engine and an electrical machine. The main components of HEVs are energy storage system, motor, bidirectional converter and maximum power point trackers (MPPT, in case of solar-powered HEVs). The performance of HEVs greatly depends on these components and its architecture. This paper presents an extensive review on essential components used in HEVs such as their architectures with advantages and disadvantages, choice of bidirectional converter to obtain high efficiency, combining ultracapacitor with battery to extend the battery life, traction motors’ role and their suitability for a particular application. Inclusion of photovoltaic cell in HEVs is a fairly new concept and has been discussed in detail. Various MPPT techniques used for solar-driven HEVs are also discussed in this paper with their suitability.展开更多
The recent development of synthesis processes to assemble graphene sheets into porous three-dimensional （3D）macroscopic structures are reviewed, including our efforts on 3D graphene structures. Mechanisms for buildi...The recent development of synthesis processes to assemble graphene sheets into porous three-dimensional （3D）macroscopic structures are reviewed, including our efforts on 3D graphene structures. Mechanisms for building 3D graphene architectures and their composite materials are also summarized. The functional systems based on 3D graphene architectures provide a significant enhancement in the efficacy due to their unique structures and properties.展开更多
Maintenance and management of genetic diversity of farm animal genetic resources （AnGR） is very important for biological, socioeconomical and cultural significance. The core concern of conservation for farm AnGR is ...Maintenance and management of genetic diversity of farm animal genetic resources （AnGR） is very important for biological, socioeconomical and cultural significance. The core concern of conservation for farm AnGR is the retention of genetic diversity of conserved populations in a long-term perspective. However, numerous factors may affect evolution of genetic diversity of a conserved population. Among those factors, the genetic architecture of conserved populations is little considered in current conservation strategies. In this study, we investigated the dynamic changes of genetic diversity of conserved populations with two scenarios on initial genetic architectures by computer simulation in which thirty polymorphic microsatellite loci were chosen to represent genetic architecture of the populations with observed heterozygosity （Ho） and expected heterozygosity （He）, observed and mean effective number of alleles （Ao and Ae）, number of polymorphic loci （NP） and the percentage of polymorphic loci （PP）, number of rare alleles （RA） and number of non-rich polymorphic loci （NRP） as the estimates of genetic diversity. The two scenarios on genetic architecture were taken into account, namely, one conserved population with same allele frequency （AS） and another one with actual allele frequency （AA）. The results showed that the magnitude of loss of genetic diversity is associated with genetic architecture of initial conserved population, the amplitude of genetic diversity decline in the context AS was more narrow extent than those in context AA, the ranges of decline of Ho and Ao were about 4 and 2 times in AA compared with that in AS, respectively, the occurrence of first monomorphic locus and the time of change of measure NP in scenario AA is 20 generations and 23 generations earlier than that in scenario AS, respectively. Additionally, we found that NRP, a novel measure proposed by our research group, was a proper estimate for monitoring the evolution of genetic diversity in a closed conserved population. Our study suggested that current managements of conserved populations should emphasize on initial genetic architecture in order to make an effective and feasible conservation scheme.展开更多
Field Programmable Gate Array(FPGA) is an efficient reconfigurable integrated circuit platform and has become a core signal processing microchip device of digital systems over the last decade. With the rapid developme...Field Programmable Gate Array(FPGA) is an efficient reconfigurable integrated circuit platform and has become a core signal processing microchip device of digital systems over the last decade. With the rapid development of semiconductor technology, the performance and system integration of FPGA devices have been significantly progressed, and at the same time new challenges arise. The design of FPGA architecture is required to evolve to meet these challenges, while also taking advantage of ever increased microchip density. This survey reviews the recent development of advanced FPGA architectures, including improvement of the programming technologies, logic blocks, interconnects, and embedded resources. Moreover, some important emerging design issues of FPGA architectures, such as novel memory based FPGAs and 3D FPGAs, are also presented to provide an outlook for future FPGA development.展开更多
Lithium-sulfur batteries(LSBs)are one of the most promising energy storage devices because of their high theoretical energy density;however,inherent issues including poor electrical conductivity and severe dissolution...Lithium-sulfur batteries(LSBs)are one of the most promising energy storage devices because of their high theoretical energy density;however,inherent issues including poor electrical conductivity and severe dissolution of S and its discharged products hinder their practical applications.MXenes have metallic conductivity,ultra-thin two-dimensional(2D)structures,rich surface functional groups,and macrostructural adjustability and have been widely used to design advanced sulfur hosts.3D network structures assembled by 2D MXene nanosheets have shown superior performance for improving reaction kinetics,accommodating and dispersing sulfur at the micro-/nanoscale,and capturing polysulfides due to their porous interconnected structure.Herein,the applications of MXene architectures related to 2D layered structures,3D multilayered structures,and 3D spherical structures as sulfur hosts are reviewed.The structure-performance relationship,challenges for current designs,and opportunities for future 3D architectures for LSBs are also analyzed.展开更多
Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully ap...Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully applied to solve many problems in academia and in industry.Although the explosion of big data applications is driving the development of ML,it also imposes severe challenges of data processing speed and scalability on conventional computer systems.Computing platforms that are dedicatedly designed for AI applications have been considered,ranging from a complement to von Neumann platforms to a“must-have”and stand-alone technical solution.These platforms,which belong to a larger category named“domain-specific computing,”focus on specific customization for AI.In this article,we focus on summarizing the recent advances in accelerator designs for deep neural networks(DNNs)-that is,DNN accelerators.We discuss various architectures that support DNN executions in terms of computing units,dataflow optimization,targeted network topologies,architectures on emerging technologies,and accelerators for emerging applications.We also provide our visions on the future trend of AI chip designs.展开更多
Self-assembly of Zn（NO3）2·6H2O and 2,3,5,6-tetrabromoterephthalic acid（H2TBTA） gave rise to two new zinc metal-organic frameworks,Zn（HTBTA）2（phen）2·H2O（1） and Zn（TBTA）1/2（μ2-OH）（H2O）·0....Self-assembly of Zn（NO3）2·6H2O and 2,3,5,6-tetrabromoterephthalic acid（H2TBTA） gave rise to two new zinc metal-organic frameworks,Zn（HTBTA）2（phen）2·H2O（1） and Zn（TBTA）1/2（μ2-OH）（H2O）·0.25EtOH（2）.Complex 1 is a mononuclear molecule.The hydrogen bonding interactions further connect the mononuclear molecules to generate a 2D supramolecular architecture.Complex 2 is a 2D organic-inorganic hybrid layer framework constructed from 1D rod-shaped secondary building units.展开更多
The photocatalytic degradation of norfloxacin by bismuth tungstate(Bi2WO6)with different hierarchical architectures wasinvestigated under visible light irradiation.Bi2WO6was prepared by hydrothermal method with the re...The photocatalytic degradation of norfloxacin by bismuth tungstate(Bi2WO6)with different hierarchical architectures wasinvestigated under visible light irradiation.Bi2WO6was prepared by hydrothermal method with the reaction solution pH rangingfrom4to11.The relatively ultrathin Bi2WO6nanoflakes prepared at pH4showed excellent adsorption and photodegradationefficiency towards norfloxacin.The characterization results showed that Bi2WO6prepared at pH4had a larger specific area andfaster photo-generated carrier separation rate.The decay rate reached the maximum in weak alkaline reaction solution,which couldbe attributed to the presence of moderate OH-anions.The present study demonstrated that the smaller size of Bi2WO6could be anefficient photocatalyst on the degradation of norfloxacin in the aquatic environment.展开更多
Three-dimensional（3D） flowerlike hierarchical Zn2GeO4 and Mn2＋-doped Zn2GeO4 microstructures have been prepared by a facile hydrothermal approach. X-Ray diffraction（XRD）, field emission scanning electron micro-sc...Three-dimensional（3D） flowerlike hierarchical Zn2GeO4 and Mn2＋-doped Zn2GeO4 microstructures have been prepared by a facile hydrothermal approach. X-Ray diffraction（XRD）, field emission scanning electron micro-scopy（FESEM）, transmission electron microscopy（TEM） and photoluminescence（PL） spectrometry were employed to characterize the samples. Such flowerlike hierarchical Zn2GeO4 microstructures with an average diameter of 3―4 μm were found to be constructed by abundant single crystalline nanorods of about 90 nm in diameter. The luminescent properties of Zn2GeO4：xMn phosphors with different contents of Mn2＋ as an activator were investigated. The Mn2＋-doped samples showed green luminescence corresponding to the d-d transition of Mn2＋ under the irradiation of UV light. The red shift（from 531 nm to 538 nm） in λem with increasing Mn2＋ content was observed in the luminescent spectra, which should be attributed to a weak crystal field because of the substitution of Zn2＋ by Mn2＋ at a distorted tetrahedral lattice site.展开更多
A new type of implantable drug delivery devices （ DDD ） with complicated architectures were fubricated by three-dimensional printing technique, employing levofloxacin （LVFX） as a model drug. Processing parameters...A new type of implantable drug delivery devices （ DDD ） with complicated architectures were fubricated by three-dimensional printing technique, employing levofloxacin （LVFX） as a model drug. Processing parameters were optimized in riew of the layer thickness, spucing between printed lines, flow rate of liquid binder and the fast axis speed. The prepared DDD prototype consists of a double-layer structure, of which the upper region is a reservoir system and the lower region is a matrix one. The in vitro release test revealed that LVFX was released in a dual-puse pattern. This DDD may present a new strategy for the prophylaxis and treatment of diseases such as bone infection in the near future.展开更多
Lithium?ion batteries(LIBs), which are high?energy?density and low?safety?risk secondary batteries, are underpinned to the rise in electrochemical energy storage devices that satisfy the urgent demands of the global e...Lithium?ion batteries(LIBs), which are high?energy?density and low?safety?risk secondary batteries, are underpinned to the rise in electrochemical energy storage devices that satisfy the urgent demands of the global energy storage market. With the aim of achiev?ing high energy density and fast?charging performance, the exploitation of simple and low?cost approaches for the production of high capacity, high density, high mass loading, and kinetically ion?accessible electrodes that maximize charge storage and transport in LIBs, is a critical need. Toward the construction of high?performance electrodes, carbons are promisingly used in the enhanced roles of active materials, electrochemi?cal reaction frameworks for high?capacity noncarbons, and lightweight current collectors. Here, we review recent advances in the carbon engi?neering of electrodes for excellent electrochemical performance and structural stability, which is enabled by assembled carbon architectures that guarantee su cient charge delivery and volume fluctuation bu ering inside the electrode during cycling. Some specific feasible assem?bly methods, synergism between structural design components of carbon assemblies, and electrochemical performance enhancement are highlighted. The precise design of carbon cages by the assembly of graphene units is potentially useful for the controlled preparation of high?capacity carbon?caged noncarbon anodes with volumetric capacities over 2100 mAh cm^(-3). Finally, insights are given on the prospects and challenges for designing carbon architectures for practical LIBs that simultaneously provide high energy densities(both gravimetric and volumetric) and high rate performance.展开更多
There are complex and regular changes on sedimentary facies from the Early to the Middle Triassic in the Nanpanjiang basin. After the obvious drowned event of carbonate platforms in the transitional period between Per...There are complex and regular changes on sedimentary facies from the Early to the Middle Triassic in the Nanpanjiang basin. After the obvious drowned event of carbonate platforms in the transitional period between Permian and Triassic, carbonate platforms have evolved into the ramp type from the rimmed shelf type. The differentiation of sedimentary facies becomes clearer in space, which are marked by the changes from an attached platform to a turbidity basin and several isolated platforms in the basin. The striking characteristics are the development of oolitic banks on isolated platforms in Nanning and Jingxi and the reef and bank limestones in the margin of the attached platform in the Early Triassic. Despite the difference of the time span and the architecture of facies succession of third order sedimentary sequences, the process of the third order relative sea level changes reflected by the sedimentary facies succession of the third order sequences is generally synchronous. Therefore, six third order sequences could be discerned in the strata from the Early to the Middle Triassic in the Nanpanjiang basin. Using two types of facies changing surfaces and two types of diachronisms in stratigraphic records as the key elements, the sedimentary facies architectures of the third order sequences that represent sequence stratigraphic frameworks from the Early to the Middle Triassic in the Nanpanjiang basin could be constructed.展开更多
We propose a content-based parallel image retrieval system to achieve high responding ability. Our system is developed on cluster architectures. It has several retrieval. servers to supply the service of content-based...We propose a content-based parallel image retrieval system to achieve high responding ability. Our system is developed on cluster architectures. It has several retrieval. servers to supply the service of content-based image retrieval. It adopts the Browser/Server (B/S) mode. The users could visit our system though web pages. It uses the symmetrical color-spatial features (SCSF) to represent the content of an image. The SCSF is effective and efficient for image matching because it is independent of image distortion such as rotation and flip as well as it increases the matching accuracy. The SCSF was organized by M-tree, which could speedup the searching procedure. Our experiments show that the image matching is quickly and efficiently with the use of SCSF. And with the support of several retrieval servers, the system could respond to many users at mean time. Key words content-based image retrieval - cluster architecture - color-spatial feature - B/S mode - task parallel - WWW - Internet CLC number TP391 Foundation item: Supported by the National Natural Science Foundation of China (60173058)Biography: ZHOU Bing (1975-), male, Ph. D candidate, reseach direction: data mining, content-based image retrieval.展开更多
Novel centralized base station architectures integrating computation and communication functionalities have become important for the development of future mobile communication networks.Therefore,the development of dyn...Novel centralized base station architectures integrating computation and communication functionalities have become important for the development of future mobile communication networks.Therefore,the development of dynamic high-speed interconnections between baseband units(BBUs)and remote radio heads(RRHs)is vital in centralized base station design.Herein,dynamic high-speed switches(HSSs)connecting BBUs and RRHs were designed for a centralized base station architecture.We analyzed the characteristics of actual traffic and introduced a switch traffic model suitable for the super base station architecture.Then,we proposed a data-priority-aware(DPA)scheduling algorithm based on the traffic model.Lastly,we developed the dynamic HSS model based on the OPNET platform and the prototype based on FPGA.Our results show that the DPA achieves close to 100%throughput with lower latency and provides better run-time complexity than iOCF and HE-iSLIP,thereby demonstrating that the proposed switch system can be adopted in centralized base station architectures.展开更多
文摘This paper introduces a computational cognitive architecture that serves as a comprehensive computational theory of the human mind,from cognitive science and computational psychology.The cognitive architecture(named Clarion)has been justified by,and validated against,psychological data,findings,and theoretical constructs.One important theoretical background for it is the dual-process theories,which led to its overall two-level structuring in a hybrid neuro-symbolic way.Furthermore,given the recent advances in AI and computing technology,LLMs are being incorporated into the model to better capture human intuition and instinct(and implicit processes in general),in order to further enhance Clarion.Integrating Clarion and LLMs can also help to develop AI systems that are more capable,more reliable,and more human-like.Overall,the paper advocates a multidisciplinary approach towards developing better models for cognitive science and for AI.
文摘A complete examination of Large Language Models’strengths,problems,and applications is needed due to their rising use across disciplines.Current studies frequently focus on single-use situations and lack a comprehensive understanding of LLM architectural performance,strengths,and weaknesses.This gap precludes finding the appropriate models for task-specific applications and limits awareness of emerging LLM optimization and deployment strategies.In this research,50 studies on 25+LLMs,including GPT-3,GPT-4,Claude 3.5,DeepKet,and hybrid multimodal frameworks like ContextDET and GeoRSCLIP,are thoroughly reviewed.We propose LLM application taxonomy by grouping techniques by task focus—healthcare,chemistry,sentiment analysis,agent-based simulations,and multimodal integration.Advanced methods like parameter-efficient tuning(LoRA),quantumenhanced embeddings(DeepKet),retrieval-augmented generation(RAG),and safety-focused models(GalaxyGPT)are evaluated for dataset requirements,computational efficiency,and performance measures.Frameworks for ethical issues,data limited hallucinations,and KDGI-enhanced fine-tuning like Woodpecker’s post-remedy corrections are highlighted.The investigation’s scope,mad,and methods are described,but the primary results are not.The work reveals that domain-specialized fine-tuned LLMs employing RAG and quantum-enhanced embeddings performbetter for context-heavy applications.In medical text normalization,ChatGPT-4 outperforms previous models,while two multimodal frameworks,GeoRSCLIP,increase remote sensing.Parameter-efficient tuning technologies like LoRA have minimal computing cost and similar performance,demonstrating the necessity for adaptive models in multiple domains.To discover the optimum domain-specific models,explain domain-specific fine-tuning,and present quantum andmultimodal LLMs to address scalability and cross-domain issues.The framework helps academics and practitioners identify,adapt,and innovate LLMs for different purposes.This work advances the field of efficient,interpretable,and ethical LLM application research.
基金the National Natural Science Foundation of China(No.62304020)supported by the National Key R&D Program of China(No.2023YFB3811300)the National Natural Science Foundation of China(No.52202370).
文摘The widespread proliferation of modern wireless devices coupled with overlapping power emissions has brought about electromagnetic(EM)pollution issues,posing many challenges to environment and human health.Therefore,the development of EM shielding devices with high green shielding index(gs)is essential,as they offer absorption-dominant protection that minimizes reflections and safeguards both health and electronics.MXene,with its intrinsic ultra-high electrical conductivity,liquid-phase tunable surface chemistry,low density,large specific surface area,thermal stability,and mechanical stability,has become the leading two-dimensional(2D)material driving the development of green EM shielding devices.In this review we emphasize device-level strategies with engineered architectures for MXene-based green EM shielding.We first examine MXene’s crystal and electronic structure and the fundamental attenuation mechanisms in MXene-based devices.Then we survey fabrication and assembly methods,analyzing three device-level strategies for MXene-based green EM shielded devices:3D architectures,metastructure/meta-surfaces,and external stimulus.Throughout,we highlight how MXene’s distinguished properties enable green EM interference(EMI)shielding devices that minimize secondary interference.Finally,we discuss the challenges faced in the effective utilization of MXene-based in green EM shielding devices,provide insights into these challenges,and offer guidelines for developing the solutions of next-generation green MXene-based EM shielding devices.
基金Supported by the Natural Science Foundation of China(61076019)the China Postdoctoral Science Foundation(20100481134)+1 种基金the Natural Science Foundation of Jiangsu Province(BK2008387)the Graduate Student Innovation Foundation of Jiangsu Province(CX07B-105z)~~
文摘The network on chip（NoC）is used as a solution for the communication problems in a complex system on chip（SoC）design.To further enhance performances,the NoC architectures,a high level modeling and an evaluation method based on OPNET are proposed to analyze their performances on different injection rates and traffic patterns.Simulation results for general NoC in terms of the average latency and the throughput are analyzed and used as a guideline to make appropriate choices for a given application.Finally,a MPEG4 decoder is mapped on different NoC architectures.Results prove the effectiveness of the evaluation method.
文摘In this paper we discuss policy iteration methods for approximate solution of a finite-state discounted Markov decision problem, with a focus on feature-based aggregation methods and their connection with deep reinforcement learning schemes. We introduce features of the states of the original problem, and we formulate a smaller "aggregate" Markov decision problem, whose states relate to the features. We discuss properties and possible implementations of this type of aggregation, including a new approach to approximate policy iteration. In this approach the policy improvement operation combines feature-based aggregation with feature construction using deep neural networks or other calculations. We argue that the cost function of a policy may be approximated much more accurately by the nonlinear function of the features provided by aggregation, than by the linear function of the features provided by neural networkbased reinforcement learning, thereby potentially leading to more effective policy improvement.
文摘The rapid consumption of fossil fuel and increased environmental damage caused by it have given a strong impetus to the growth and development of fuelefficient vehicles. Hybrid electric vehicles (HEVs) have evolved from their inchoate state and are proving to be a promising solution to the serious existential problem posed to the planet earth. Not only do HEVs provide better fuel economy and lower emissions satisfying environmental legislations, but also they dampen the effect of rising fuel prices on consumers. HEVs combine the drive powers of an internal combustion engine and an electrical machine. The main components of HEVs are energy storage system, motor, bidirectional converter and maximum power point trackers (MPPT, in case of solar-powered HEVs). The performance of HEVs greatly depends on these components and its architecture. This paper presents an extensive review on essential components used in HEVs such as their architectures with advantages and disadvantages, choice of bidirectional converter to obtain high efficiency, combining ultracapacitor with battery to extend the battery life, traction motors’ role and their suitability for a particular application. Inclusion of photovoltaic cell in HEVs is a fairly new concept and has been discussed in detail. Various MPPT techniques used for solar-driven HEVs are also discussed in this paper with their suitability.
基金supported by the National Natural Science Foundation of China (Grant Nos. 11004230, 51172273, 11290161, and 11027402)the National Basic Research Program of China (Grant Nos. 2012CB933003 and 2013CB932603)the Chinese Academy of Sciences (Grant No. KJCX2-YW-W35)
文摘The recent development of synthesis processes to assemble graphene sheets into porous three-dimensional （3D）macroscopic structures are reviewed, including our efforts on 3D graphene structures. Mechanisms for building 3D graphene architectures and their composite materials are also summarized. The functional systems based on 3D graphene architectures provide a significant enhancement in the efficacy due to their unique structures and properties.
基金partly supported by the program for Changjiang Scholar and Innovation Research Team in University, China (IRT1191)the Beijing Science and Technology Project, China (Z080005022208015)the Chinese Universities Scientific Fund (2012YJ058)
文摘Maintenance and management of genetic diversity of farm animal genetic resources （AnGR） is very important for biological, socioeconomical and cultural significance. The core concern of conservation for farm AnGR is the retention of genetic diversity of conserved populations in a long-term perspective. However, numerous factors may affect evolution of genetic diversity of a conserved population. Among those factors, the genetic architecture of conserved populations is little considered in current conservation strategies. In this study, we investigated the dynamic changes of genetic diversity of conserved populations with two scenarios on initial genetic architectures by computer simulation in which thirty polymorphic microsatellite loci were chosen to represent genetic architecture of the populations with observed heterozygosity （Ho） and expected heterozygosity （He）, observed and mean effective number of alleles （Ao and Ae）, number of polymorphic loci （NP） and the percentage of polymorphic loci （PP）, number of rare alleles （RA） and number of non-rich polymorphic loci （NRP） as the estimates of genetic diversity. The two scenarios on genetic architecture were taken into account, namely, one conserved population with same allele frequency （AS） and another one with actual allele frequency （AA）. The results showed that the magnitude of loss of genetic diversity is associated with genetic architecture of initial conserved population, the amplitude of genetic diversity decline in the context AS was more narrow extent than those in context AA, the ranges of decline of Ho and Ao were about 4 and 2 times in AA compared with that in AS, respectively, the occurrence of first monomorphic locus and the time of change of measure NP in scenario AA is 20 generations and 23 generations earlier than that in scenario AS, respectively. Additionally, we found that NRP, a novel measure proposed by our research group, was a proper estimate for monitoring the evolution of genetic diversity in a closed conserved population. Our study suggested that current managements of conserved populations should emphasize on initial genetic architecture in order to make an effective and feasible conservation scheme.
基金Supported by National Natural Science Foundation of China(No.61271149)National High Technology Research and Development Program of China(No.2012AA-012301)National Science and Technology Major Project of China(No.2013ZX03006004)
文摘Field Programmable Gate Array(FPGA) is an efficient reconfigurable integrated circuit platform and has become a core signal processing microchip device of digital systems over the last decade. With the rapid development of semiconductor technology, the performance and system integration of FPGA devices have been significantly progressed, and at the same time new challenges arise. The design of FPGA architecture is required to evolve to meet these challenges, while also taking advantage of ever increased microchip density. This survey reviews the recent development of advanced FPGA architectures, including improvement of the programming technologies, logic blocks, interconnects, and embedded resources. Moreover, some important emerging design issues of FPGA architectures, such as novel memory based FPGAs and 3D FPGAs, are also presented to provide an outlook for future FPGA development.
基金supported by the National Natural Science Foundation of China(21805105,21975091 and 21773078)。
文摘Lithium-sulfur batteries(LSBs)are one of the most promising energy storage devices because of their high theoretical energy density;however,inherent issues including poor electrical conductivity and severe dissolution of S and its discharged products hinder their practical applications.MXenes have metallic conductivity,ultra-thin two-dimensional(2D)structures,rich surface functional groups,and macrostructural adjustability and have been widely used to design advanced sulfur hosts.3D network structures assembled by 2D MXene nanosheets have shown superior performance for improving reaction kinetics,accommodating and dispersing sulfur at the micro-/nanoscale,and capturing polysulfides due to their porous interconnected structure.Herein,the applications of MXene architectures related to 2D layered structures,3D multilayered structures,and 3D spherical structures as sulfur hosts are reviewed.The structure-performance relationship,challenges for current designs,and opportunities for future 3D architectures for LSBs are also analyzed.
基金the National Science Foundations(NSFs)(1822085,1725456,1816833,1500848,1719160,and 1725447)the NSF Computing and Communication Foundations(1740352)+1 种基金the Nanoelectronics COmputing REsearch Program in the Semiconductor Research Corporation(NC-2766-A)the Center for Research in Intelligent Storage and Processing-in-Memory,one of six centers in the Joint University Microelectronics Program,a SRC program sponsored by Defense Advanced Research Projects Agency.
文摘Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully applied to solve many problems in academia and in industry.Although the explosion of big data applications is driving the development of ML,it also imposes severe challenges of data processing speed and scalability on conventional computer systems.Computing platforms that are dedicatedly designed for AI applications have been considered,ranging from a complement to von Neumann platforms to a“must-have”and stand-alone technical solution.These platforms,which belong to a larger category named“domain-specific computing,”focus on specific customization for AI.In this article,we focus on summarizing the recent advances in accelerator designs for deep neural networks(DNNs)-that is,DNN accelerators.We discuss various architectures that support DNN executions in terms of computing units,dataflow optimization,targeted network topologies,architectures on emerging technologies,and accelerators for emerging applications.We also provide our visions on the future trend of AI chip designs.
基金Supported by the NNSFC (20701025)the NSF of Shandong Province (Y2008B01),and Shandong University
文摘Self-assembly of Zn（NO3）2·6H2O and 2,3,5,6-tetrabromoterephthalic acid（H2TBTA） gave rise to two new zinc metal-organic frameworks,Zn（HTBTA）2（phen）2·H2O（1） and Zn（TBTA）1/2（μ2-OH）（H2O）·0.25EtOH（2）.Complex 1 is a mononuclear molecule.The hydrogen bonding interactions further connect the mononuclear molecules to generate a 2D supramolecular architecture.Complex 2 is a 2D organic-inorganic hybrid layer framework constructed from 1D rod-shaped secondary building units.
基金Projects(51579096,51222805,51521006,51508175) supported by the National Natural Science Foundation of ChinaProject supported by the National Program for Support of Top-Notch Young Professionals of China+1 种基金Project(NCET–11–0129) supported by the Program for New Century Excellent Talents in University from the Ministry of Education of ChinaProject(CX2015B095) supported by the Hunan Province Innovation Foundation for Postgraduate,China
文摘The photocatalytic degradation of norfloxacin by bismuth tungstate(Bi2WO6)with different hierarchical architectures wasinvestigated under visible light irradiation.Bi2WO6was prepared by hydrothermal method with the reaction solution pH rangingfrom4to11.The relatively ultrathin Bi2WO6nanoflakes prepared at pH4showed excellent adsorption and photodegradationefficiency towards norfloxacin.The characterization results showed that Bi2WO6prepared at pH4had a larger specific area andfaster photo-generated carrier separation rate.The decay rate reached the maximum in weak alkaline reaction solution,which couldbe attributed to the presence of moderate OH-anions.The present study demonstrated that the smaller size of Bi2WO6could be anefficient photocatalyst on the degradation of norfloxacin in the aquatic environment.
基金Supported by the National Natural Science Foundation of China(No.21073032)
文摘Three-dimensional（3D） flowerlike hierarchical Zn2GeO4 and Mn2＋-doped Zn2GeO4 microstructures have been prepared by a facile hydrothermal approach. X-Ray diffraction（XRD）, field emission scanning electron micro-scopy（FESEM）, transmission electron microscopy（TEM） and photoluminescence（PL） spectrometry were employed to characterize the samples. Such flowerlike hierarchical Zn2GeO4 microstructures with an average diameter of 3―4 μm were found to be constructed by abundant single crystalline nanorods of about 90 nm in diameter. The luminescent properties of Zn2GeO4：xMn phosphors with different contents of Mn2＋ as an activator were investigated. The Mn2＋-doped samples showed green luminescence corresponding to the d-d transition of Mn2＋ under the irradiation of UV light. The red shift（from 531 nm to 538 nm） in λem with increasing Mn2＋ content was observed in the luminescent spectra, which should be attributed to a weak crystal field because of the substitution of Zn2＋ by Mn2＋ at a distorted tetrahedral lattice site.
文摘A new type of implantable drug delivery devices （ DDD ） with complicated architectures were fubricated by three-dimensional printing technique, employing levofloxacin （LVFX） as a model drug. Processing parameters were optimized in riew of the layer thickness, spucing between printed lines, flow rate of liquid binder and the fast axis speed. The prepared DDD prototype consists of a double-layer structure, of which the upper region is a reservoir system and the lower region is a matrix one. The in vitro release test revealed that LVFX was released in a dual-puse pattern. This DDD may present a new strategy for the prophylaxis and treatment of diseases such as bone infection in the near future.
基金supported by the National Science Fund for Distinguished Young Scholars of China (No. 51525204)National Key Basic Research Program of China (2014CB932400)the National Natural Science Foundation of China (No. 51872195 and U1401243)
文摘Lithium?ion batteries(LIBs), which are high?energy?density and low?safety?risk secondary batteries, are underpinned to the rise in electrochemical energy storage devices that satisfy the urgent demands of the global energy storage market. With the aim of achiev?ing high energy density and fast?charging performance, the exploitation of simple and low?cost approaches for the production of high capacity, high density, high mass loading, and kinetically ion?accessible electrodes that maximize charge storage and transport in LIBs, is a critical need. Toward the construction of high?performance electrodes, carbons are promisingly used in the enhanced roles of active materials, electrochemi?cal reaction frameworks for high?capacity noncarbons, and lightweight current collectors. Here, we review recent advances in the carbon engi?neering of electrodes for excellent electrochemical performance and structural stability, which is enabled by assembled carbon architectures that guarantee su cient charge delivery and volume fluctuation bu ering inside the electrode during cycling. Some specific feasible assem?bly methods, synergism between structural design components of carbon assemblies, and electrochemical performance enhancement are highlighted. The precise design of carbon cages by the assembly of graphene units is potentially useful for the controlled preparation of high?capacity carbon?caged noncarbon anodes with volumetric capacities over 2100 mAh cm^(-3). Finally, insights are given on the prospects and challenges for designing carbon architectures for practical LIBs that simultaneously provide high energy densities(both gravimetric and volumetric) and high rate performance.
基金ThestudyisjointlysupportedbytheChinaPetroleumCorporation (No .NPJ- 10 0 19)andalsobytheMinistryofScienceandTechnology (SSER)
文摘There are complex and regular changes on sedimentary facies from the Early to the Middle Triassic in the Nanpanjiang basin. After the obvious drowned event of carbonate platforms in the transitional period between Permian and Triassic, carbonate platforms have evolved into the ramp type from the rimmed shelf type. The differentiation of sedimentary facies becomes clearer in space, which are marked by the changes from an attached platform to a turbidity basin and several isolated platforms in the basin. The striking characteristics are the development of oolitic banks on isolated platforms in Nanning and Jingxi and the reef and bank limestones in the margin of the attached platform in the Early Triassic. Despite the difference of the time span and the architecture of facies succession of third order sedimentary sequences, the process of the third order relative sea level changes reflected by the sedimentary facies succession of the third order sequences is generally synchronous. Therefore, six third order sequences could be discerned in the strata from the Early to the Middle Triassic in the Nanpanjiang basin. Using two types of facies changing surfaces and two types of diachronisms in stratigraphic records as the key elements, the sedimentary facies architectures of the third order sequences that represent sequence stratigraphic frameworks from the Early to the Middle Triassic in the Nanpanjiang basin could be constructed.
文摘We propose a content-based parallel image retrieval system to achieve high responding ability. Our system is developed on cluster architectures. It has several retrieval. servers to supply the service of content-based image retrieval. It adopts the Browser/Server (B/S) mode. The users could visit our system though web pages. It uses the symmetrical color-spatial features (SCSF) to represent the content of an image. The SCSF is effective and efficient for image matching because it is independent of image distortion such as rotation and flip as well as it increases the matching accuracy. The SCSF was organized by M-tree, which could speedup the searching procedure. Our experiments show that the image matching is quickly and efficiently with the use of SCSF. And with the support of several retrieval servers, the system could respond to many users at mean time. Key words content-based image retrieval - cluster architecture - color-spatial feature - B/S mode - task parallel - WWW - Internet CLC number TP391 Foundation item: Supported by the National Natural Science Foundation of China (60173058)Biography: ZHOU Bing (1975-), male, Ph. D candidate, reseach direction: data mining, content-based image retrieval.
基金the key project of the National Science and Technology Major Project(Grant No.2018ZX03001017)the project of the CAS engineering laboratory for intelligent agricultural machinery equipment(Grant No.GC201907-02).
文摘Novel centralized base station architectures integrating computation and communication functionalities have become important for the development of future mobile communication networks.Therefore,the development of dynamic high-speed interconnections between baseband units(BBUs)and remote radio heads(RRHs)is vital in centralized base station design.Herein,dynamic high-speed switches(HSSs)connecting BBUs and RRHs were designed for a centralized base station architecture.We analyzed the characteristics of actual traffic and introduced a switch traffic model suitable for the super base station architecture.Then,we proposed a data-priority-aware(DPA)scheduling algorithm based on the traffic model.Lastly,we developed the dynamic HSS model based on the OPNET platform and the prototype based on FPGA.Our results show that the DPA achieves close to 100%throughput with lower latency and provides better run-time complexity than iOCF and HE-iSLIP,thereby demonstrating that the proposed switch system can be adopted in centralized base station architectures.
