Vision Transformers (ViTs) have become a universal backbone for both image recognition and image generation. Yet their Multi–Head Self–Attention (MHSA) layer still performs a quadratic query–key interaction for \emph{every} token pair, spending the bulk of computation on visually weak or redundant correlations. We introduce \emph{Visual–Contrast Attention} (VCA), a drop-in replacement for MHSA that injects an explicit notion of discrimination while reducing the theoretical complexity from $\mathcal{O}(N^{2}C)$ to $\mathcal{O}(N n C)$ with $n\!\ll\!N$. VCA first distils each head’s dense query field into a handful of spatially pooled \emph{visual–contrast tokens}, then splits them into a learnable \emph{positive} and \emph{negative} stream whose differential interaction highlights what truly separates one region from another. The module adds fewer than $0.3$\,M parameters to a DeiT-Tiny backbone, requires no extra FLOPs, and is wholly architecture-agnostic. Empirically, VCA lifts DeiT-Tiny top-1 accuracy on ImageNet-1K from $72.2\%$ to \textbf{$75.6\%$} (+$3.4$) and improves three strong hierarchical ViTs by up to $3.1$\%, while in class-conditional ImageNet generation it lowers FID-50K by $2.1$ to $5.2$ points across both diffusion (DiT) and flow (SiT) models. Extensive ablations confirm that (i) spatial pooling supplies low-variance global cues, (ii) dual positional embeddings are indispensable for contrastive reasoning, and (iii) combining the two in both stages yields the strongest synergy. VCA therefore offers a simple path towards faster and sharper Vision Transformers. The source code is available at \href{https://github.com/LeapLabTHU/LinearDiff}{https://github.com/LeapLabTHU/LinearDiff}.

Remaining useful life (RUL) prediction of aircraft engine (AE) is of great importance to improve its reliability and availability, and reduce its maintenance costs. This article proposes a novel deep bidirectional recurrent neural networks (DBRNNs) ensemble method for the RUL prediction of the AEs. In this method, several kinds of DBRNNs with different neuron structures are built to extract hidden features from sensory data. A new customized loss function is designed to evaluate the performance of the DBRNNs, and a series of the RUL values is obtained. Then, these RUL values are reencapsulated into a predicted RUL domain. By updating the weights of elements in the domain, multiple regression decision tree (RDT) models are trained iteratively. These models integrate the predicted results of different DBRNNs to realize the final RUL prognostics with high accuracy. The proposed method is validated by using C-MAPSS datasets from NASA. The experimental results show that the proposed method has achieved more superior performance compared with other existing methods.

Discrete Gaussian sampling is one of the important components in lattice-based cryptosystems which are promising candidates for post-quantum cryptographic algorithms. For sufficient security and satisfactory performance, the Knuth-Yao algorithm is an efficient way to implement discrete Gaussian samplers. Nevertheless, most polynomials in lattice-based cryptography have 256 coefficients or more, which suffers from long latency to complete the sample generation. In this paper, the first parallel discrete Gaussian sampler with hierarchical structure is proposed, while keeping statistical distance to the actual distribution. Based on the imbalanced visiting frequency of the probability matrix, a three-stage generation strategy is adopted with hierarchical bit search units (BSUs) that can greatly reduce area consumption of the repeated costly lookup tables. Besides the architecture improvement, a lowest-set-bit scanning scheme is introduced to BSUs. Moreover, the parallelism of our design provides obfuscation ability against side-channel attacks (SCAs). A practical hardware implementation of discrete Gaussian distributions with $\sigma$=3.33 on the Xilinx Virtex-5 XC5VLX30 FPGA device spends 26.12 ns on average to generate 256 samples, consuming 994 slices. Results have verified its advantages of area efficiency over the state-of-the-arts (SOAs).

Machine health monitoring is of great importance in industrial informatics field. Recently, deep learning methods applied to machine health monitoring have been proven effective. However, the existing methods face enormous difficulties in extracting heterogeneous features indicating the variation until failure and revealing the inherent high-dimensional features of massive signals, which affect the accuracy and efficiency of machine health monitoring. In this paper, a novel data-driven machine health monitoring method is proposed using adaptive kernel spectral clustering (AKSC) and deep long short-term memory recurrent neural networks (LSTM-RNN). This method include three steps: First, features in the time domain, frequency domain, and time-frequency domain are, respectively, extracted from massive measured signals. And, an Euclidean distance based algorithm is designed to select degradation features. Second, the AKSC algorithm is introduced to adaptively identify machine anomaly behaviors from multiple degradation features. Third, a new deep learning model (LSTM-RNN) is constructed to update and predict the failure time of the machine. The effectiveness of the proposed method is validated using a set of test-to-failure experimental data. The results show that the performance of the proposed method is competitive with other existing methods.

A mixed-model two-sided assembly line is a manufacturing system designed for the production of large-sized products. In order to describe the actual condition, this paper presents a novel multiobjective programming model for balancing a mixed-model two-sided assembly line subject to multiple constraints, in which, additional constraints including zoning, synchronous, and positional constraints are considered besides the traditional constraints, e.g., the precedence constraint. Two objectives are simultaneously to be optimized, one is to minimize the combination of the weighted line efficiency and the weighted smoothness index, and the other is to minimize the weighted total relevant costs per unit of a product. A novel multiobjective hybrid imperialist competitive algorithm (MOHICA) is proposed to solve this problem. In the presented MOHICA, the sigma method is employed to quantify every individual, a novel merging method is introduced to reserve better individuals into the evolutionary population, and late acceptance hill-climbing (LAHC) algorithm is presented as a local search algorithm to achieve accurate balance between intensification and diversification. The experimental results on the selected benchmark instances and a practical case show that the proposed multiobjective algorithm outperforms nondominated sorting genetic algorithm (NSGA)-II, multiobjective improved teaching-learning-based optimization, and NSGA-III existing in the literature.

Ship design is a complex process, which contains many coupled disciplines and needs the collaboration among geographical distributed design resources. Multidisciplinary Design Optimization (MDO) provides a promising way to ship design compared to traditional spiral design approach. As the important enabling tool, MDO framework is vital to the successful implementation of MDO problems. In this paper, at the base of surveying the state of the art of some notable MDO frameworks, a Web services and Isight based distributed ship MDO framework is proposed, which provides an easy way to integrate distributed design, analysis, optimization services, namely various tools and programs, to jointly implement ship design. At the same time, according to the requirements of ship design, some new and extended function modules are added, which incorporates a number of existing technologies such as workflow, component technology.

Analytical Target Cascading (ATC) is a product development tool that computes component design specifications so that the final system design is consistent and meets design targets. It is a promising methodology for decomposition of complex system design and optimization. However, ATC has some shortcomings inherently. In this paper, after analyzing the theory and model of ATC, a more promising method entitled Analytical Target Cascading Based on Physical Programming (ATC-PP) is proposed. This new approach is intended to substantially place the design process into a more flexible and practical framework by adding the advantages of Physical Programming (PP) into ATC. And the preference function which is proposed in PP is used to definite the deviation between response and linking variables in ATC in the new method. It can completely eliminate the need for assigning weight coefficient, which is the object of the typical computational bottleneck in large design optimization problems. And it could also give the designers' preferences in practical design. Finally, the validity of ATC-PP is approved by studying an anchor design problem.

Traditionally, process planning and scheduling were performed sequentially, where scheduling was done after process plans had been generated. Considering the fact that the two functions are usually complementary, it is necessary to integrate them more tightly so that the performance of a manufacturing system can be improved greatly. In this paper, a Multi-agent-based approach has been developed to facilitate the integration of the two functions. In the approach, the two functions are carried out simultaneously, and an optimization agent based on an evolutionary algorithm is used to manage the interactions and communications between agents to enable proper decisions to be made. To verify the feasibility and performance of the proposed approach, an experimental study has been conducted and comparisons have been made between this approach and some previous works. The experimental results show the proposed approach has achieved significant improvement.

Collaborative design is a complex process being carried out by distributed teams. Collaboration requires successful and efficient sharing of knowledge, negotiation, coordination and management of activities. In distributed environments, design members are moving to network forms through alliances and other collaborative relationships as a social network. This paper explores collaborative design process from a sociological viewpoint. A social factor-based architecture for collaborative design is investigated, with special attention given to patterns of social behavior and the organizations of those who participate in the design. Social factors thought to have played a role include perspectives, preference, attitudes, cognition, etc. The main contribution of this paper is to introduce a methodology for analyzing the social aspects of collaborative design. The paper concludes with directing attention to current challenges and recommendations for research.

A test-bed (HUST-CDMS) for collaborative product design and manufacturing is described in this paper. This test-bed is constructed by using a CORBA and MAS based system architecture, where diverse and heterogeneous design and manufacturing resources are MAS-based agentified and CORBA-based encapsulated as resource agents and subsequently integrated in the system. These resource agents communicate and interact with each other by means of agent interoperation and human-machine interaction to achieve a collaborative design and manufacturing environment, aiming at reducing the lead time and costs

Collaborative design can be characterized by different level constraints that describe basic elements of design process such as human, information and resource. In design, satisfied solutions to constraints must occur to arrive at a final design result. Few collaborative design systems exist for adequately representing constraints, and formally combining them from different disciplines. The main contribution of this paper is to introduce a methodology for analyzing the constraints in collaborative design. A hierarchy of fuzzy constraint networks is presented along with a methodology for creating interlinks between different levels of the hierarchy. The solution methodology is illustrated with an example. The paper concludes with directing attention to current challenges and recommendations for research

At present, the research of workflow modeling has been extended to the context of virtual enterprise in order to realize business process integration among member enterprises. Petri net and its extensions are classic modeling methods in workflow modeling, while the process-view is a promising approach used in workflow composition. This paper proposes a novel Petri net based process-view method to model workflow for virtual enterprise and concentrates on the mapping from the Petri net based base workflow model to its corresponding process-views, which incorporates the coordination of control flow and data flow. During the mapping process, abstract place and logical transition are proposed and added in order to abstract information from the base workflow model and also to ensure that the process-views are complete Petri net workflow models. At last, an integrated workflow model can be obtained by integrating the process-views, which is the interface to realize cross-organizational business process integration