The Underwater Acoustic(UWA)channel is bandwidth-constrained and experiences doubly selective fading.It is challenging to acquire perfect channel knowledge for Orthogonal Frequency Division Multiplexing(OFDM)communica...The Underwater Acoustic(UWA)channel is bandwidth-constrained and experiences doubly selective fading.It is challenging to acquire perfect channel knowledge for Orthogonal Frequency Division Multiplexing(OFDM)communications using a finite number of pilots.On the other hand,Deep Learning(DL)approaches have been very successful in wireless OFDM communications.However,whether they will work underwater is still a mystery.For the first time,this paper compares two categories of DL-based UWA OFDM receivers:the DataDriven(DD)method,which performs as an end-to-end black box,and the Model-Driven(MD)method,also known as the model-based data-driven method,which combines DL and expert OFDM receiver knowledge.The encoder-decoder framework and Convolutional Neural Network(CNN)structure are employed to establish the DD receiver.On the other hand,an unfolding-based Minimum Mean Square Error(MMSE)structure is adopted for the MD receiver.We analyze the characteristics of different receivers by Monte Carlo simulations under diverse communications conditions and propose a strategy for selecting a proper receiver under different communication scenarios.Field trials in the pool and sea are also conducted to verify the feasibility and advantages of the DL receivers.It is observed that DL receivers perform better than conventional receivers in terms of bit error rate.展开更多
Under complex working conditions,accurate prediction of the remaining useful life(RUL)of lithium-ion batteries is of great significance to ensure the stable operation of energy storage systems,the safe driving of elec...Under complex working conditions,accurate prediction of the remaining useful life(RUL)of lithium-ion batteries is of great significance to ensure the stable operation of energy storage systems,the safe driving of electric vehicles,and the continuous power supply of electronic devices.This paper systematically describes the RUL prediction methods of lithium-ion batteries and comprehensively summarizes the development status and future trends in this field.First,the battery degradation mechanisms and lightweight data acquisition are analyzed.Secondly,a systematic classification model is constructed for the more widely used lithium battery RUL prediction methods,and the application characteristics and implementation limitations of different methods are analyzed in detail.An innovative classification framework for hybrid methods is proposed based on the depth of physical-data interaction.Then,collaborative modelling of calendar ageing and cyclic ageing is discussed,revealing their coupled effects and corresponding RUL prediction methods.Finally,the technical bottlenecks faced by the current RUL prediction of lithium batteries are identified,potential solutions are proposed,and the future development trends are outlined.展开更多
Salient object detection(SOD)models struggle to simultaneously preserve global structure,maintain sharp object boundaries,and sustain computational efficiency in complex scenes.In this study,we propose SPSALNet,a task...Salient object detection(SOD)models struggle to simultaneously preserve global structure,maintain sharp object boundaries,and sustain computational efficiency in complex scenes.In this study,we propose SPSALNet,a task-driven two-stage(macro–micro)architecture that restructures the SOD process around superpixel representations.In the proposed approach,a“split-and-enhance”principle,introduced to our knowledge for the first time in the SOD literature,hierarchically classifies superpixels and then applies targeted refinement only to ambiguous or error-prone regions.At the macro stage,the image is partitioned into content-adaptive superpixel regions,and each superpixel is represented by a high-dimensional region-level feature vector.These representations define a regional decomposition problem in which superpixels are assigned to three classes:background,object interior,and transition regions.Superpixel tokens interact with a global feature vector from a deep network backbone through a cross-attention module and are projected into an enriched embedding space that jointly encodes local topology and global context.At the micro stage,the model employs a U-Net-based refinement process that allocates computational resources only to ambiguous transition regions.The image and distance–similarity maps derived from superpixels are processed through a dual-encoder pathway.Subsequently,channel-aware fusion blocks adaptively combine information from these two sources,producing sharper and more stable object boundaries.Experimental results show that SPSALNet achieves high accuracy with lower computational cost compared to recent competing methods.On the PASCAL-S and DUT-OMRON datasets,SPSALNet exhibits a clear performance advantage across all key metrics,and it ranks first on accuracy-oriented measures on HKU-IS.On the challenging DUT-OMRON benchmark,SPSALNet reaches a MAE of 0.034.Across all datasets,it preserves object boundaries and regional structure in a stable and competitive manner.展开更多
Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challeng...Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.展开更多
Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training dataset...Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training datasets limit the training performance and generalization.The model-driven deep learning introduces the diffraction model into the neural network.It eliminates the need for the labeled training dataset and has been extensively applied to hologram generation.However,the existing model-driven deep learning algorithms face the problem of insufficient constraints.In this study,we propose a model-driven neural network capable of high-fidelity 4K computer-generated hologram generation,called 4K Diffraction Model-driven Network(4K-DMDNet).The constraint of the reconstructed images in the frequency domain is strengthened.And a network structure that combines the residual method and sub-pixel convolution method is built,which effectively enhances the fitting ability of the network for inverse problems.The generalization of the 4K-DMDNet is demonstrated with binary,grayscale and 3D images.High-quality full-color optical reconstructions of the 4K holograms have been achieved at the wavelengths of 450 nm,520 nm,and 638 nm.展开更多
Model-Driven Engineering (MDE) by reframing software development as the transformation of high-level models, promises lots of gains to Software Engineering in terms of productivity, quality and reusability. Although a...Model-Driven Engineering (MDE) by reframing software development as the transformation of high-level models, promises lots of gains to Software Engineering in terms of productivity, quality and reusability. Although a number of empirical studies have established the reality of these gains, there are still lots of reluctances toward the adoption of MDE in practice. This resistance can be explained by several technological and social factors among which a natural scepticism toward novel approaches. In this paper we attempt to provide arguments to help alleviate this scepticism by conducting an assessment of a MDE approach. Our goal is to show that although this MDE is novel, it retains similarities with the conventional Software Engineering approach while automating aspects of it.展开更多
Although the Model-Driven paradigm is being accepted in the research environment as a very useful and powerful option for effective software development, its real application in the enterprise context is still a chall...Although the Model-Driven paradigm is being accepted in the research environment as a very useful and powerful option for effective software development, its real application in the enterprise context is still a challenge for software engineering. Several causes can be stacked out, but one of them can be the lack of tool support for the efficient application of this paradigm. This paper presents a set of tools, grouped in a suite named NDT-Suite, which under the Model-Driven paradigm offer a suitable solution for software development. These tools explore different options that this paradigm can improve such as, development, quality assurance or requirement treatment. Besides, this paper analyses how they are being successfully applied in the industry.展开更多
In a context where urban satellite image processing technologies are undergoing rapid evolution,this article presents an innovative and rigorous approach to satellite image classification applied to urban planning.Thi...In a context where urban satellite image processing technologies are undergoing rapid evolution,this article presents an innovative and rigorous approach to satellite image classification applied to urban planning.This research proposes an integrated methodological framework,based on the principles of model-driven engineering(MDE),to transform a generic meta-model into a meta-model specifically dedicated to urban satellite image classification.We implemented this transformation using the Atlas Transformation Language(ATL),guaranteeing a smooth and consistent transition from platform-independent model(PIM)to platform-specific model(PSM),according to the principles of model-driven architecture(MDA).The application of this IDM methodology enables advanced structuring of satellite data for targeted urban planning analyses,making it possible to classify various urban zones such as built-up,cultivated,arid and water areas.The novelty of this approach lies in the automation and standardization of the classification process,which significantly reduces the need for manual intervention,and thus improves the reliability,reproducibility and efficiency of urban data analysis.By adopting this method,decision-makers and urban planners are provided with a powerful tool for systematically and consistently analyzing and interpreting satellite images,facilitating decision-making in critical areas such as urban space management,infrastructure planning and environmental preservation.展开更多
In the context of reducing its carbon emissions,the Chinese steel industry is currently undergoing an intelligent transformation to enhance its profitability and sustainability.The optimization of production planning ...In the context of reducing its carbon emissions,the Chinese steel industry is currently undergoing an intelligent transformation to enhance its profitability and sustainability.The optimization of production planning and scheduling plays a pivotal role in realizing these objectives such as improving production efficiency,saving energy,reducing carbon emissions,and enhancing quality.However,current practices in steel enterprises are largely dependent on experience-driven manual decision approaches supported by information systems,which are inadequate to meet the complex requirements of the industry.This study explores the current situation in production planning and scheduling,analyzes the characteristics and limitations of existing methods,and emphasizes the necessity and trends of intelligent systems.It surveys the current literature on production planning and scheduling in steel enterprises and analyzes the theoretical advancements and practical challenges associated with combinatorial and sequential optimization in this field.A key focus is on the limitations of current models and algorithms in effectively addressing the multi-objective and multiconstraint characteristics of steel produc-tion.To overcome these challenges,a novel framework for intelligent production planning and scheduling is proposed.This framework leverages data-and knowledge-driven decision-making and scenario adaptability,enabling the system to respond dynamically to real-time production conditions and market fluctuations.By integrating artificial intelligence and advanced optimization methodologies,the proposed framework improves the efficiency,cost-effectiveness,and environmental sustainability of steel manufacturing.展开更多
This study explored the transformative potential of artificial intelligence(AI)in addressing the challenges posed by terahertz ultra-massive multiple-input multiple-output(UM-MIMO)systems.It begins by outlining the ch...This study explored the transformative potential of artificial intelligence(AI)in addressing the challenges posed by terahertz ultra-massive multiple-input multiple-output(UM-MIMO)systems.It begins by outlining the characteristics of terahertz UM-MIMO systems and identifies three primary challenges for transceiver design:computational complexity,modeling difficulty,and measurement limitations.The study posits that AI provides a promising solution to these challenges.Three systematic research roadmaps are proposed for developing AI algorithms tailored to terahertz UM-MIMO systems.The first roadmap,model-driven deep learning(DL),emphasizes the importance of leveraging available domain knowledge and advocates the adoption of AI only to enhance bottleneck modules within an established signal processing or optimization framework.Four essential steps are discussed:algorithmic frameworks,basis algorithms,loss function design,and neural architecture design.The second roadmap presents channel state information(CSI)foundation models,aimed at unifying the design of different transceiver modules by focusing on their shared foundation,that is,the wireless channel.The training of a single compact foundation model is proposed to estimate the score function of wireless channels,which serve as a versatile prior for designing a wide variety of transceiver modules.Four essential steps are outlined:general frameworks,conditioning,site-specific adaptation,and the joint design of CSI foundation models and model-driven DL.The third roadmap aims to explore potential directions for applying pretrained large language models(LLMs)to terahertz UM-MIMO systems.Several application scenarios are envisioned,including LLM-based estimation,optimization,search,network management,and protocol understanding.Finally,the study highlights open problems and future research directions.展开更多
To reduce complexity, the combat effectiveness simulation system(CESS) is often decomposed into static structure,physical behavior, and cognitive behavior, and model abstraction is layered onto domain invariant knowle...To reduce complexity, the combat effectiveness simulation system(CESS) is often decomposed into static structure,physical behavior, and cognitive behavior, and model abstraction is layered onto domain invariant knowledge(DIK) and application variant knowledge(AVK) levels. This study concentrates on the specification of CESS’s physical behaviors at the DIK level of abstraction, and proposes a model driven framework for efficiently developing simulation models within model-driven engineering(MDE). Technically, this framework integrates the four-layer metamodeling architecture and a set of model transformation techniques with the objective of reducing model heterogeneity and enhancing model continuity. As a proof of concept, a torpedo example is illustrated to explain how physical models are developed following the proposed framework. Finally, a combat scenario is constructed to demonstrate the availability, and a further verification is shown by a reasonable agreement between simulation results and field observations.展开更多
Software testing has been attracting a lot of attention for effective software development.In model driven approach,Unified Modelling Language(UML)is a conceptual modelling approach for obligations and other features ...Software testing has been attracting a lot of attention for effective software development.In model driven approach,Unified Modelling Language(UML)is a conceptual modelling approach for obligations and other features of the system in a model-driven methodology.Specialized tools interpret these models into other software artifacts such as code,test data and documentation.The generation of test cases permits the appropriate test data to be determined that have the aptitude to ascertain the requirements.This paper focuses on optimizing the test data obtained from UML activity and state chart diagrams by using Basic Genetic Algorithm(BGA).For generating the test cases,both diagrams were converted into their corresponding intermediate graphical forms namely,Activity Diagram Graph(ADG)and State Chart Diagram Graph(SCDG).Then both graphs will be combined to form a single graph called,Activity State Chart Diagram Graph(ASCDG).Both graphs were then joined to create a single graph known as the Activity State Chart Diagram Graph(ASCDG).Next,the ASCDG will be optimized using BGA to generate the test data.A case study involving a withdrawal from the automated teller machine(ATM)of a bank was employed to demonstrate the approach.The approach successfully identified defects in various ATM functions such as messaging and operation.展开更多
Recently,the ontological metamodel plays an increasingly important role to specify systems in two forms:ontology and metamodel.Ontology is a descriptive model representing reality by a set of concepts,their interrelat...Recently,the ontological metamodel plays an increasingly important role to specify systems in two forms:ontology and metamodel.Ontology is a descriptive model representing reality by a set of concepts,their interrelations,and constraints.On the other hand,metamodel is a more classical,but more powerful model in which concepts and relationships are represented in a prescriptive way.This study firstly clarifies the difference between the two approaches,then explains their advantages and limitations,and attempts to explore a general ontological metamodeling framework by integrating each characteristic,in order to implement semantic simulation model engineering.As a proof of concept,this paper takes the combat effectiveness simulation systems as a motivating case,uses the proposed framework to define a set of ontological composable modeling frameworks,and presents an underwater targets search scenario for running simulations and analyzing results.Finally,this paper expects that this framework will be generally used in other fields.展开更多
The production capacity of shale oil reservoirs after hydraulic fracturing is influenced by a complex interplay involving geological characteristics,engineering quality,and well conditions.These relationships,nonlinea...The production capacity of shale oil reservoirs after hydraulic fracturing is influenced by a complex interplay involving geological characteristics,engineering quality,and well conditions.These relationships,nonlinear in nature,pose challenges for accurate description through physical models.While field data provides insights into real-world effects,its limited volume and quality restrict its utility.Complementing this,numerical simulation models offer effective support.To harness the strengths of both data-driven and model-driven approaches,this study established a shale oil production capacity prediction model based on a machine learning combination model.Leveraging fracturing development data from 236 wells in the field,a data-driven method employing the random forest algorithm is implemented to identify the main controlling factors for different types of shale oil reservoirs.Through the combination model integrating support vector machine(SVM)algorithm and back propagation neural network(BPNN),a model-driven shale oil production capacity prediction model is developed,capable of swiftly responding to shale oil development performance under varying geological,fluid,and well conditions.The results of numerical experiments show that the proposed method demonstrates a notable enhancement in R2 by 22.5%and 5.8%compared to singular machine learning models like SVM and BPNN,showcasing its superior precision in predicting shale oil production capacity across diverse datasets.展开更多
Combat system effectiveness simulation (CSES) is a special type of complex system simulation. Three non-functional requirements (NFRs), i.e. model composability, domain specific modeling, and model evolvability, are g...Combat system effectiveness simulation (CSES) is a special type of complex system simulation. Three non-functional requirements (NFRs), i.e. model composability, domain specific modeling, and model evolvability, are gaining higher priority from CSES users when evaluating different modeling methodologies for CSES. Traditional CSES modeling methodologies are either domain-neutral (lack of domain characteristics consideration and limited support for model composability) or domain-oriented (lack of openness and evolvability) and fall short of the three NFRs. Inspired by the concept of architecture in systems engineering and software engineering fields, we extend it into a concept of model architecture for complex simulation systems, and propose a model architecture-oriented modeling methodology in which the model architecture plays a central role in achieving the three NFRs. Various model-driven engineering (MDE) approaches and technologies, including simulation modeling platform (SMP), unified modeling language (UML), domain specific modeling (DSM), eclipse modeling framework (EMF), graphical modeling framework (GMF), and so forth, are applied where possible in representing the CSES model architecture and its components' behaviors from physical and cognitive domain aspects. A prototype CSES system, called weapon effectiveness simulation system (WESS), and a non-trivial air-combat simulation example are presented to demonstrate the methodology.展开更多
Testing in Software Engineering is one of the most important phases although, unfortunately, it cannot be always successfully fulfilled due to time constraints. In most cases, the development phase takes more time tha...Testing in Software Engineering is one of the most important phases although, unfortunately, it cannot be always successfully fulfilled due to time constraints. In most cases, the development phase takes more time than it was estimated, entailing negative effects on the testing phase. The delay increases even more in Research and Development (R + D) projects, where the real time to execute tasks is more difficult to control. Model Driven Engineering (MDE) offers a solution to avoid testing costs without affecting the execution quality of the applied test. This paper presents a practical overview of a Model Driven Testing (MDT)-based methodology and its impact on CALIPSOneo project, which was carried out in liaison with Airbus Defense and Space and, particularly, with the Product Lifecycle Management (PLM) department.展开更多
This work presents the design of an Internet of Things(IoT)edge-based system based on model transformation and complete weighted graph to detect violations of social distancing measures in indoor public places.Awirele...This work presents the design of an Internet of Things(IoT)edge-based system based on model transformation and complete weighted graph to detect violations of social distancing measures in indoor public places.Awireless sensor network based on Bluetooth Low Energy is introduced as the infrastructure of the proposed design.A hybrid model transformation strategy for generating a graph database to represent groups of people is presented as a core middleware layer of the detecting system’s proposed architectural design.A Neo4j graph database is used as a target implementation generated from the proposed transformational system to store all captured real-time IoT data about the distances between individuals in an indoor area and answer user predefined queries,expressed using Neo4j Cypher,to provide insights from the stored data for decision support.As proof of concept,a discrete-time simulation model was adopted for the design of a COVID-19 physical distancing measures case study to evaluate the introduced system architecture.Twenty-one weighted graphs were generated randomly and the degrees of violation of distancing measures were inspected.The experimental results demonstrate the capability of the proposed system design to detect violations of COVID-19 physical distancing measures within an enclosed area.展开更多
Recent estimates indicate that more than half the software market belongs to enterprise applications. One of the greatest challenges in these is in conducting the complex process of adaptation of pre-packaged applicat...Recent estimates indicate that more than half the software market belongs to enterprise applications. One of the greatest challenges in these is in conducting the complex process of adaptation of pre-packaged applications, such as Oracle or SAP, to the organization needs. Although very detailed, structured and well documented methods govern this process, the consulting team implementing the method must spend much manual effort in making sure that the guidelines of the method are followed as intended by the method author. The problem is exacerbated by the diversity of skills and roles of team members, and the many sorts of communications of collaboration that methods prescribe. By enhancing the metamodel in which the methods are defined, we automatically produce a CASE tool (so to speak) for the applications of these methods. Our results are successfully employed in a number of large, ongoing projects with demonstrable, non-meager saving.展开更多
The emerging Model-Driven Engineering (MDE) paradigm advocates the use of models as first-class citizens in the software development process, while artifacts such as documentation and source-code can be quickly produc...The emerging Model-Driven Engineering (MDE) paradigm advocates the use of models as first-class citizens in the software development process, while artifacts such as documentation and source-code can be quickly produced from those models by using automated transformations. Even though many MDE-oriented approaches, languages and tools have been developed in the recent past, there is no standard that concretely defines a specific sequence of steps to obtain a functional software system from a model. Thus, the existing approaches present numerous differences among themselves, because each one handles the problems inherent to software development in its own way. This paper presents and discusses a reference model for the comparative study of current MDE approaches in the scope of web-application development. This reference model focuses on relevant aspects such as modeling language scope (domain, business-logic, user-interface), usage of patterns, separation of concerns, model transformations, tool support, and deployment details like web-platform independence and traditional programming required. The ultimate goal of this paper is to determine the aspects that will be of greater importance in future web-oriented MDE languages.展开更多
基金funded in part by the National Natural Science Foundation of China under Grant 62401167 and 62192712in part by the Key Laboratory of Marine Environmental Survey Technology and Application,Ministry of Natural Resources,P.R.China under Grant MESTA-2023-B001in part by the Stable Supporting Fund of National Key Laboratory of Underwater Acoustic Technology under Grant JCKYS2022604SSJS007.
文摘The Underwater Acoustic(UWA)channel is bandwidth-constrained and experiences doubly selective fading.It is challenging to acquire perfect channel knowledge for Orthogonal Frequency Division Multiplexing(OFDM)communications using a finite number of pilots.On the other hand,Deep Learning(DL)approaches have been very successful in wireless OFDM communications.However,whether they will work underwater is still a mystery.For the first time,this paper compares two categories of DL-based UWA OFDM receivers:the DataDriven(DD)method,which performs as an end-to-end black box,and the Model-Driven(MD)method,also known as the model-based data-driven method,which combines DL and expert OFDM receiver knowledge.The encoder-decoder framework and Convolutional Neural Network(CNN)structure are employed to establish the DD receiver.On the other hand,an unfolding-based Minimum Mean Square Error(MMSE)structure is adopted for the MD receiver.We analyze the characteristics of different receivers by Monte Carlo simulations under diverse communications conditions and propose a strategy for selecting a proper receiver under different communication scenarios.Field trials in the pool and sea are also conducted to verify the feasibility and advantages of the DL receivers.It is observed that DL receivers perform better than conventional receivers in terms of bit error rate.
基金supported by the National Natural Science Foundation of China(No.U23A20651)the Central Government Guides Local Science and Technology Development Foundation(No.2023ZYDF022)+1 种基金the Sichuan Science and Technology Program(2024ZDZX0031)the Open Fund Project of State Key Laboratory of Mining Response and Disaster Prevention and Control in Deep Coal Mines(No.SKLMRDPC23KF19).
文摘Under complex working conditions,accurate prediction of the remaining useful life(RUL)of lithium-ion batteries is of great significance to ensure the stable operation of energy storage systems,the safe driving of electric vehicles,and the continuous power supply of electronic devices.This paper systematically describes the RUL prediction methods of lithium-ion batteries and comprehensively summarizes the development status and future trends in this field.First,the battery degradation mechanisms and lightweight data acquisition are analyzed.Secondly,a systematic classification model is constructed for the more widely used lithium battery RUL prediction methods,and the application characteristics and implementation limitations of different methods are analyzed in detail.An innovative classification framework for hybrid methods is proposed based on the depth of physical-data interaction.Then,collaborative modelling of calendar ageing and cyclic ageing is discussed,revealing their coupled effects and corresponding RUL prediction methods.Finally,the technical bottlenecks faced by the current RUL prediction of lithium batteries are identified,potential solutions are proposed,and the future development trends are outlined.
文摘Salient object detection(SOD)models struggle to simultaneously preserve global structure,maintain sharp object boundaries,and sustain computational efficiency in complex scenes.In this study,we propose SPSALNet,a task-driven two-stage(macro–micro)architecture that restructures the SOD process around superpixel representations.In the proposed approach,a“split-and-enhance”principle,introduced to our knowledge for the first time in the SOD literature,hierarchically classifies superpixels and then applies targeted refinement only to ambiguous or error-prone regions.At the macro stage,the image is partitioned into content-adaptive superpixel regions,and each superpixel is represented by a high-dimensional region-level feature vector.These representations define a regional decomposition problem in which superpixels are assigned to three classes:background,object interior,and transition regions.Superpixel tokens interact with a global feature vector from a deep network backbone through a cross-attention module and are projected into an enriched embedding space that jointly encodes local topology and global context.At the micro stage,the model employs a U-Net-based refinement process that allocates computational resources only to ambiguous transition regions.The image and distance–similarity maps derived from superpixels are processed through a dual-encoder pathway.Subsequently,channel-aware fusion blocks adaptively combine information from these two sources,producing sharper and more stable object boundaries.Experimental results show that SPSALNet achieves high accuracy with lower computational cost compared to recent competing methods.On the PASCAL-S and DUT-OMRON datasets,SPSALNet exhibits a clear performance advantage across all key metrics,and it ranks first on accuracy-oriented measures on HKU-IS.On the challenging DUT-OMRON benchmark,SPSALNet reaches a MAE of 0.034.Across all datasets,it preserves object boundaries and regional structure in a stable and competitive manner.
文摘Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.
基金We are grateful for financial supports from National Natural Science Foundation of China(62035003,61775117)China Postdoctoral Science Foundation(BX2021140)Tsinghua University Initiative Scientific Research Program(20193080075).
文摘Deep learning offers a novel opportunity to achieve both high-quality and high-speed computer-generated holography(CGH).Current data-driven deep learning algorithms face the challenge that the labeled training datasets limit the training performance and generalization.The model-driven deep learning introduces the diffraction model into the neural network.It eliminates the need for the labeled training dataset and has been extensively applied to hologram generation.However,the existing model-driven deep learning algorithms face the problem of insufficient constraints.In this study,we propose a model-driven neural network capable of high-fidelity 4K computer-generated hologram generation,called 4K Diffraction Model-driven Network(4K-DMDNet).The constraint of the reconstructed images in the frequency domain is strengthened.And a network structure that combines the residual method and sub-pixel convolution method is built,which effectively enhances the fitting ability of the network for inverse problems.The generalization of the 4K-DMDNet is demonstrated with binary,grayscale and 3D images.High-quality full-color optical reconstructions of the 4K holograms have been achieved at the wavelengths of 450 nm,520 nm,and 638 nm.
文摘Model-Driven Engineering (MDE) by reframing software development as the transformation of high-level models, promises lots of gains to Software Engineering in terms of productivity, quality and reusability. Although a number of empirical studies have established the reality of these gains, there are still lots of reluctances toward the adoption of MDE in practice. This resistance can be explained by several technological and social factors among which a natural scepticism toward novel approaches. In this paper we attempt to provide arguments to help alleviate this scepticism by conducting an assessment of a MDE approach. Our goal is to show that although this MDE is novel, it retains similarities with the conventional Software Engineering approach while automating aspects of it.
文摘Although the Model-Driven paradigm is being accepted in the research environment as a very useful and powerful option for effective software development, its real application in the enterprise context is still a challenge for software engineering. Several causes can be stacked out, but one of them can be the lack of tool support for the efficient application of this paradigm. This paper presents a set of tools, grouped in a suite named NDT-Suite, which under the Model-Driven paradigm offer a suitable solution for software development. These tools explore different options that this paradigm can improve such as, development, quality assurance or requirement treatment. Besides, this paper analyses how they are being successfully applied in the industry.
文摘In a context where urban satellite image processing technologies are undergoing rapid evolution,this article presents an innovative and rigorous approach to satellite image classification applied to urban planning.This research proposes an integrated methodological framework,based on the principles of model-driven engineering(MDE),to transform a generic meta-model into a meta-model specifically dedicated to urban satellite image classification.We implemented this transformation using the Atlas Transformation Language(ATL),guaranteeing a smooth and consistent transition from platform-independent model(PIM)to platform-specific model(PSM),according to the principles of model-driven architecture(MDA).The application of this IDM methodology enables advanced structuring of satellite data for targeted urban planning analyses,making it possible to classify various urban zones such as built-up,cultivated,arid and water areas.The novelty of this approach lies in the automation and standardization of the classification process,which significantly reduces the need for manual intervention,and thus improves the reliability,reproducibility and efficiency of urban data analysis.By adopting this method,decision-makers and urban planners are provided with a powerful tool for systematically and consistently analyzing and interpreting satellite images,facilitating decision-making in critical areas such as urban space management,infrastructure planning and environmental preservation.
基金supported by the Key Program of the National Natural Science Foundation of China(Nos.52334008 and 51734004).
文摘In the context of reducing its carbon emissions,the Chinese steel industry is currently undergoing an intelligent transformation to enhance its profitability and sustainability.The optimization of production planning and scheduling plays a pivotal role in realizing these objectives such as improving production efficiency,saving energy,reducing carbon emissions,and enhancing quality.However,current practices in steel enterprises are largely dependent on experience-driven manual decision approaches supported by information systems,which are inadequate to meet the complex requirements of the industry.This study explores the current situation in production planning and scheduling,analyzes the characteristics and limitations of existing methods,and emphasizes the necessity and trends of intelligent systems.It surveys the current literature on production planning and scheduling in steel enterprises and analyzes the theoretical advancements and practical challenges associated with combinatorial and sequential optimization in this field.A key focus is on the limitations of current models and algorithms in effectively addressing the multi-objective and multiconstraint characteristics of steel produc-tion.To overcome these challenges,a novel framework for intelligent production planning and scheduling is proposed.This framework leverages data-and knowledge-driven decision-making and scenario adaptability,enabling the system to respond dynamically to real-time production conditions and market fluctuations.By integrating artificial intelligence and advanced optimization methodologies,the proposed framework improves the efficiency,cost-effectiveness,and environmental sustainability of steel manufacturing.
基金supported in part by the Hong Kong Research Grant Council(16209023)。
文摘This study explored the transformative potential of artificial intelligence(AI)in addressing the challenges posed by terahertz ultra-massive multiple-input multiple-output(UM-MIMO)systems.It begins by outlining the characteristics of terahertz UM-MIMO systems and identifies three primary challenges for transceiver design:computational complexity,modeling difficulty,and measurement limitations.The study posits that AI provides a promising solution to these challenges.Three systematic research roadmaps are proposed for developing AI algorithms tailored to terahertz UM-MIMO systems.The first roadmap,model-driven deep learning(DL),emphasizes the importance of leveraging available domain knowledge and advocates the adoption of AI only to enhance bottleneck modules within an established signal processing or optimization framework.Four essential steps are discussed:algorithmic frameworks,basis algorithms,loss function design,and neural architecture design.The second roadmap presents channel state information(CSI)foundation models,aimed at unifying the design of different transceiver modules by focusing on their shared foundation,that is,the wireless channel.The training of a single compact foundation model is proposed to estimate the score function of wireless channels,which serve as a versatile prior for designing a wide variety of transceiver modules.Four essential steps are outlined:general frameworks,conditioning,site-specific adaptation,and the joint design of CSI foundation models and model-driven DL.The third roadmap aims to explore potential directions for applying pretrained large language models(LLMs)to terahertz UM-MIMO systems.Several application scenarios are envisioned,including LLM-based estimation,optimization,search,network management,and protocol understanding.Finally,the study highlights open problems and future research directions.
基金supported by the National Natural Science Foundation of China(61273198)
文摘To reduce complexity, the combat effectiveness simulation system(CESS) is often decomposed into static structure,physical behavior, and cognitive behavior, and model abstraction is layered onto domain invariant knowledge(DIK) and application variant knowledge(AVK) levels. This study concentrates on the specification of CESS’s physical behaviors at the DIK level of abstraction, and proposes a model driven framework for efficiently developing simulation models within model-driven engineering(MDE). Technically, this framework integrates the four-layer metamodeling architecture and a set of model transformation techniques with the objective of reducing model heterogeneity and enhancing model continuity. As a proof of concept, a torpedo example is illustrated to explain how physical models are developed following the proposed framework. Finally, a combat scenario is constructed to demonstrate the availability, and a further verification is shown by a reasonable agreement between simulation results and field observations.
基金support from the Deanship of Scientific Research,University of Hail,Saudi Arabia through the project Ref.(RG-191315).
文摘Software testing has been attracting a lot of attention for effective software development.In model driven approach,Unified Modelling Language(UML)is a conceptual modelling approach for obligations and other features of the system in a model-driven methodology.Specialized tools interpret these models into other software artifacts such as code,test data and documentation.The generation of test cases permits the appropriate test data to be determined that have the aptitude to ascertain the requirements.This paper focuses on optimizing the test data obtained from UML activity and state chart diagrams by using Basic Genetic Algorithm(BGA).For generating the test cases,both diagrams were converted into their corresponding intermediate graphical forms namely,Activity Diagram Graph(ADG)and State Chart Diagram Graph(SCDG).Then both graphs will be combined to form a single graph called,Activity State Chart Diagram Graph(ASCDG).Both graphs were then joined to create a single graph known as the Activity State Chart Diagram Graph(ASCDG).Next,the ASCDG will be optimized using BGA to generate the test data.A case study involving a withdrawal from the automated teller machine(ATM)of a bank was employed to demonstrate the approach.The approach successfully identified defects in various ATM functions such as messaging and operation.
基金the National Natural Science Foundation of China(61273198).
文摘Recently,the ontological metamodel plays an increasingly important role to specify systems in two forms:ontology and metamodel.Ontology is a descriptive model representing reality by a set of concepts,their interrelations,and constraints.On the other hand,metamodel is a more classical,but more powerful model in which concepts and relationships are represented in a prescriptive way.This study firstly clarifies the difference between the two approaches,then explains their advantages and limitations,and attempts to explore a general ontological metamodeling framework by integrating each characteristic,in order to implement semantic simulation model engineering.As a proof of concept,this paper takes the combat effectiveness simulation systems as a motivating case,uses the proposed framework to define a set of ontological composable modeling frameworks,and presents an underwater targets search scenario for running simulations and analyzing results.Finally,this paper expects that this framework will be generally used in other fields.
基金supported by the China Postdoctoral Science Foundation(2021M702304)Natural Science Foundation of Shandong Province(ZR20210E260).
文摘The production capacity of shale oil reservoirs after hydraulic fracturing is influenced by a complex interplay involving geological characteristics,engineering quality,and well conditions.These relationships,nonlinear in nature,pose challenges for accurate description through physical models.While field data provides insights into real-world effects,its limited volume and quality restrict its utility.Complementing this,numerical simulation models offer effective support.To harness the strengths of both data-driven and model-driven approaches,this study established a shale oil production capacity prediction model based on a machine learning combination model.Leveraging fracturing development data from 236 wells in the field,a data-driven method employing the random forest algorithm is implemented to identify the main controlling factors for different types of shale oil reservoirs.Through the combination model integrating support vector machine(SVM)algorithm and back propagation neural network(BPNN),a model-driven shale oil production capacity prediction model is developed,capable of swiftly responding to shale oil development performance under varying geological,fluid,and well conditions.The results of numerical experiments show that the proposed method demonstrates a notable enhancement in R2 by 22.5%and 5.8%compared to singular machine learning models like SVM and BPNN,showcasing its superior precision in predicting shale oil production capacity across diverse datasets.
基金supported by the National Natural Science Foundation of China(61273198)
文摘Combat system effectiveness simulation (CSES) is a special type of complex system simulation. Three non-functional requirements (NFRs), i.e. model composability, domain specific modeling, and model evolvability, are gaining higher priority from CSES users when evaluating different modeling methodologies for CSES. Traditional CSES modeling methodologies are either domain-neutral (lack of domain characteristics consideration and limited support for model composability) or domain-oriented (lack of openness and evolvability) and fall short of the three NFRs. Inspired by the concept of architecture in systems engineering and software engineering fields, we extend it into a concept of model architecture for complex simulation systems, and propose a model architecture-oriented modeling methodology in which the model architecture plays a central role in achieving the three NFRs. Various model-driven engineering (MDE) approaches and technologies, including simulation modeling platform (SMP), unified modeling language (UML), domain specific modeling (DSM), eclipse modeling framework (EMF), graphical modeling framework (GMF), and so forth, are applied where possible in representing the CSES model architecture and its components' behaviors from physical and cognitive domain aspects. A prototype CSES system, called weapon effectiveness simulation system (WESS), and a non-trivial air-combat simulation example are presented to demonstrate the methodology.
文摘Testing in Software Engineering is one of the most important phases although, unfortunately, it cannot be always successfully fulfilled due to time constraints. In most cases, the development phase takes more time than it was estimated, entailing negative effects on the testing phase. The delay increases even more in Research and Development (R + D) projects, where the real time to execute tasks is more difficult to control. Model Driven Engineering (MDE) offers a solution to avoid testing costs without affecting the execution quality of the applied test. This paper presents a practical overview of a Model Driven Testing (MDT)-based methodology and its impact on CALIPSOneo project, which was carried out in liaison with Airbus Defense and Space and, particularly, with the Product Lifecycle Management (PLM) department.
文摘This work presents the design of an Internet of Things(IoT)edge-based system based on model transformation and complete weighted graph to detect violations of social distancing measures in indoor public places.Awireless sensor network based on Bluetooth Low Energy is introduced as the infrastructure of the proposed design.A hybrid model transformation strategy for generating a graph database to represent groups of people is presented as a core middleware layer of the detecting system’s proposed architectural design.A Neo4j graph database is used as a target implementation generated from the proposed transformational system to store all captured real-time IoT data about the distances between individuals in an indoor area and answer user predefined queries,expressed using Neo4j Cypher,to provide insights from the stored data for decision support.As proof of concept,a discrete-time simulation model was adopted for the design of a COVID-19 physical distancing measures case study to evaluate the introduced system architecture.Twenty-one weighted graphs were generated randomly and the degrees of violation of distancing measures were inspected.The experimental results demonstrate the capability of the proposed system design to detect violations of COVID-19 physical distancing measures within an enclosed area.
文摘Recent estimates indicate that more than half the software market belongs to enterprise applications. One of the greatest challenges in these is in conducting the complex process of adaptation of pre-packaged applications, such as Oracle or SAP, to the organization needs. Although very detailed, structured and well documented methods govern this process, the consulting team implementing the method must spend much manual effort in making sure that the guidelines of the method are followed as intended by the method author. The problem is exacerbated by the diversity of skills and roles of team members, and the many sorts of communications of collaboration that methods prescribe. By enhancing the metamodel in which the methods are defined, we automatically produce a CASE tool (so to speak) for the applications of these methods. Our results are successfully employed in a number of large, ongoing projects with demonstrable, non-meager saving.
文摘The emerging Model-Driven Engineering (MDE) paradigm advocates the use of models as first-class citizens in the software development process, while artifacts such as documentation and source-code can be quickly produced from those models by using automated transformations. Even though many MDE-oriented approaches, languages and tools have been developed in the recent past, there is no standard that concretely defines a specific sequence of steps to obtain a functional software system from a model. Thus, the existing approaches present numerous differences among themselves, because each one handles the problems inherent to software development in its own way. This paper presents and discusses a reference model for the comparative study of current MDE approaches in the scope of web-application development. This reference model focuses on relevant aspects such as modeling language scope (domain, business-logic, user-interface), usage of patterns, separation of concerns, model transformations, tool support, and deployment details like web-platform independence and traditional programming required. The ultimate goal of this paper is to determine the aspects that will be of greater importance in future web-oriented MDE languages.
