Search results

In the field of steel defect detection, the existing detection algorithms struggle to achieve a satisfactory balance between detection accuracy, computational cost and inference speed due to the interference from complex background information, the variety of defect types and significant variations in defect morphology. To solve this problem, this paper aims to propose an efficient detector based on multi-scale information extraction (MSI-YOLO), which uses YOLOv8s as the baseline model.

First, the authors introduce an efficient multi-scale convolution with different-sized convolution kernels, which enables the feature extraction network to accommodate significant variations in defect morphology. Furthermore, the authors introduce the channel prior convolutional attention mechanism, which allows the network to focus on defect areas and ignore complex background interference. Considering the lightweight design and accuracy improvement, the authors introduce a more lightweight feature fusion network (Slim-neck) to improve the fusion effect of feature maps.

MSI-YOLO achieves 79.9% mean average precision on the public data set Northeastern University (NEU)-DET, with a model size of only 19.0 MB and an frames per second of 62.5. Compared with other state-of-the-art detectors, MSI-YOLO greatly improves the recognition accuracy and has significant advantages in computational cost and inference speed. Additionally, the strong generalization ability of MSI-YOLO is verified on the collected industrial site steel data set.

This paper proposes an efficient steel defect detector with high accuracy, low computational cost, excellent detection speed and strong generalization ability, which is more valuable for practical applications in resource-limited industrial production.

Cost and schedule overruns are rife in dam projects. Normative evidence espouses overruns as though they are inimical to development and prosperity aspirations of stakeholders. This study examines the causal relationship between project financing and overruns.

Causative data were extracted from completion reports of 28 major dam projects in Africa. Each of the projects was financed jointly by up to 10 international development lenders. Relationships between causes of overruns and project outcomes were analysed.

Analyses elicit indicators of remarkable correlations between finance procedures and project outcomes. Lenders’ disposition to risk attenuation was the main debacles to project success. Interests had mounted, whilst release of fund was erratic and ill-timed. Finance objectives and mechanisms were grossly inadequate for projects’ intense bifurcations. Projects had slowed or stalled because lenders’ risks attenuation processes were purposed to favour lenders’ objectives, and not projects’ interests. In addition, findings also show project owners’ own funds and the number of lenders to a single project correlate with overruns.

Findings imply commercial complexities around major projects. They also show transactions are shaped by subtle (mis)trust behaviours in project finance procedures. Thus, scholarly solutions to project performance issues should consider behavioural issues of stakeholding parties more broadly, beyond contractors and project owners. Project finance ecosystems are vulnerable to major actors’ self-interests, opportunism and predatory conducts. Borrowers would manage this by developing and improving their capacity to build resilience and trust. Evidence shows intense borrower nations in Africa have limited capacity and acuity for these.

This study contextualises megaprojects in complexity rather than cost. Its additionality is in how finance steers absolute control of project environment away from project owners and how finance administration triggers risks and overrun.

This paper aims to provide a better understanding of the challenges and potential solutions in Visual Simultaneous Localization and Mapping (SLAM), laying the foundation for its applications in autonomous navigation, intelligent driving and other related domains.

In analyzing the latest research, the review presents representative achievements, including methods to enhance efficiency, robustness and accuracy. Additionally, the review provides insights into the future development direction of Visual SLAM, emphasizing the importance of improving system robustness when dealing with dynamic environments. The research methodology of this review involves a literature review and data set analysis, enabling a comprehensive understanding of the current status and prospects in the field of Visual SLAM.

This review aims to comprehensively evaluate the latest advances and challenges in the field of Visual SLAM. By collecting and analyzing relevant research papers and classic data sets, it reveals the current issues faced by Visual SLAM in complex environments and proposes potential solutions. The review begins by introducing the fundamental principles and application areas of Visual SLAM, followed by an in-depth discussion of the challenges encountered when dealing with dynamic objects and complex environments. To enhance the performance of SLAM algorithms, researchers have made progress by integrating different sensor modalities, improving feature extraction and incorporating deep learning techniques, driving advancements in the field.

To the best of the authors’ knowledge, the originality of this review lies in its in-depth analysis of current research hotspots and predictions for future development, providing valuable references for researchers in this field.

Accurate and timely cost prediction is critical to the success of construction projects which is still facing challenges especially at the early stage. In the context of rapid development of machine learning technology and the massive cost data from historical projects, this paper aims to propose a novel cost prediction model based on historical data with improved performance when only limited information about the new project is available.

The proposed approach combines regression analysis (RA) and artificial neural network (ANN) to build a novel hybrid cost prediction model with the former as front-end prediction and the latter as back-end correction. Firstly, the main factors influencing the cost of building projects are identified through literature research and subsequently screened by principal component analysis (PCA). Secondly the optimal RA model is determined through multi-model comparison and used for front-end prediction. Finally, ANN is applied to construct the error correction model. The hybrid RA-ANN model was trained and tested with cost data from 128 completed construction projects in China.

The results show that the hybrid cost prediction model has the advantages of both RA and ANN whose prediction accuracy is higher than that of RA and ANN only with the information such as total floor area, height and number of floors.

(1) The most critical influencing factors of the buildings’ cost are found out by means of PCA on the historical data. (2) A novel hybrid RA-ANN model is proposed which proved to have the advantages of both RA and ANN with higher accuracy. (3) The comparison among different models has been carried out which is helpful to future model selection.

The purpose of this paper is twofold: (1) to map and analyze existing complexity typologies and (2) to develop a framework for characterizing complexity-based approaches.

This study was conducted in three stages: (1) initial identification of typologies related to complexity following a structured procedure based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol; (2) backward and forward review to identify additional relevant typologies and (3) content analysis of the selected typologies, categorization and framework development.

Based on 17 selected typologies, a comprehensive overview of complexity studies is provided. Each typology is described considering key concepts, contributions and convergences and differences between them. The epistemological, theoretical and methodological diversity of complexity studies was explored, allowing the identification of the main schools of thought and authors. A framework for characterizing complexity-based approaches was proposed including the following perspectives: ontology of complexity, epistemology of complexity, purpose and object of interest, methodology and methods and theoretical pillars.

This study examines the main typologies of complexity from an integrated and multidisciplinary perspective and, based on that, proposes a novel framework to understanding and characterizing complexity-based approaches.

Predictive current control (PCC) of three-to-five-phase direct matrix converters (DMCs) is computationally expensive. For this reason, this study aims to consider a reduced number of switching states of DMC in PCC algorithm to predict the control objectives, such as output current control and input reactive power control.

The switching sequences which yield the voltage vectors of variable amplitude at a constant frequency in space are considered for the prediction and optimization step of PCC algorithm. For the selected voltage vectors, the phase angles of the output vectors are independent on the phase angles of the input vectors. In a three-to-five-phase DMC, there are 243 valid switching states. Among the switching states, only 91 states are considered using the aforementioned concept of variable amplitude output at a constant frequency. This reduced number of switching states simplifies the computational complexity of MPC based current control of three-to-five-phase DMC.

The computational complexity of the proposed PCC based DMC is lower than the all 243 vectors based PCC. The current total harmonic distortion, transient current response and input reactive power control for the simplified 91 vector based PCC are similar to the all 243 vectors based PCC.

A reduced number of switching sequence is considered for the prediction and optimization step of PCC algorithm. Hence, PCC algorithm can be sampled at a high frequency in real-time applications. Then, the performance of the PCC will be improved.

3D steady heat conduction analysis considering heat source is conducted on the fundamental of the fast multipole method (FMM)-accelerated line integration boundary element method (LIBEM).

Due to considering the heat source, domain integral is generated in the traditional heat conduction boundary integral equation (BIE), which will counteract the well-known merit of the BEM, namely, boundary-only discretization. To avoid volume discretization, the enhanced BEM, the LIBEM with dimension reduction property is introduced to transfer the domain integral into line integrals. Besides, owing to the unsatisfactory performance of the LIBEM when it comes to large-scale structures requiring massive computation, the FMM-accelerated LIBEM (FM-LIBEM) is proposed to improve the computation efficiency further.

Assuming N and M are the numbers of nodes and integral lines, respectively, the FM-LIBEM can reduce the time complexity from O(NM) to about O(N+ M), and a full discussion and verification of the advantage are done based on numerical examples under heat conduction.

(1) The LIBEM is applied to 3D heat conduction analysis with heat source. (2) The domain integrals can be transformed into boundary integrals with straight line integrals by the LIM. (3) A FM-LIBEM is proposed and can reduce the time complexity from O(NM) to O(N+ M). (4) The FM-LIBEM with high computational efficiency is exerted to solve 3D heat conduction analysis with heat source in massive computation successfully.

This paper studies the location-inventory problem (LIP) in pallet pooling systems to improve resource utilization and save logistics costs, which is a new extension of the classical LIP and also an application of the LIP in pallet pooling systems.

A mixed-integer linear programming is established, considering the location problem of pallet pooling centers (PPCs) with multi-level capacity, multi-period inventory management and bi-directional logistics. Owing to the computational complexity of the problem, a hybrid genetic algorithm (GA) is then proposed, where three local searching strategies are designed to improve the problem-solving efficiency. Lastly, numerical experiments are carried out to validate the feasibility of the established model and the efficiency of the proposed algorithm.

The results of numerical experiments show that (1) the proposed model can obtain the integrated optimal solution of the location problem and inventory management, which is better than the two-stage model and the model with single-level capacity; (2) the total cost and network structure are sensitive to the number of PPCs, the unit inventory cost, the proportion of repairable pallets and the fixed transportation cost and (3) the proposed hybrid GA shows good performance in terms of solution quality and computational time.

The established model extends the classical LIP by considering more practical factors, and the proposed algorithm provides support for solving large-scale problems. In addition, this study can also offer valuable decision support for managers in pallet pooling systems.

The purpose of this paper is to present an automatic approach for mesh sizing field generation of complicated  computer-aided design (CAD) models.

In this paper, the authors present an automatic approach for mesh sizing field generation. First, a source point extraction algorithm is applied to capture curvature and proximity features of CAD models. Second, according to the distribution of feature source points, an octree background mesh is constructed for storing element size value. Third, mesh size value on each node of background mesh is calculated by interpolating the local feature size of the nearby source points, and then, an initial mesh sizing field is obtained. Finally, a theoretically guaranteed smoothing algorithm is developed to restrict the gradient of the mesh sizing field.

To achieve high performance, the proposed approach has been implemented in multithreaded parallel using OpenMP. Numerical results demonstrate that the proposed approach is remarkably efficient to construct reasonable mesh sizing field for complicated CAD models and applicable for generating geometrically adaptive triangle/tetrahedral meshes. Moreover, since the mesh sizing field is defined on an octree background mesh, high-efficiency query of local size value could be achieved in the following mesh generation procedure.

How to determine a reasonable mesh size for complicated CAD models is often a bottleneck of mesh generation. For the complicated models with thousands or even ten thousands of geometric entities, it is time-consuming to construct an appropriate mesh sizing field for generating high-quality mesh. A parallel algorithm of mesh sizing field generation with low computational complexity is presented in this paper, and its usability and efficiency have been verified.

The anomaly detection task for oil and gas pipelines based on acoustic signals faces issues such as background noise coverage, lack of effective features, and small sample sizes, resulting in low fault identification accuracy and slow efficiency. The purpose of this paper is to study an accurate and efficient method of pipeline anomaly detection.

First, to address the impact of background noise on the accuracy of anomaly signals, the adaptive multi-threshold center frequency variational mode decomposition method(AMTCF-VMD) method is used to eliminate strong noise in pipeline signals. Secondly, to address the strong data dependency and loss of local features in the Swin Transformer network, a Hybrid Pyramid ConvNet network with an Agent Attention mechanism is proposed. This compensates for the limitations of CNN’s receptive field and enhances the Swin Transformer’s global contextual feature representation capabilities. Thirdly, to address the sparsity and imbalance of anomaly samples, the SpecAugment and Scaper methods are integrated to enhance the model’s generalization ability.

In the pipeline anomaly audio and environmental datasets such as ESC-50, the AMTCF-VMD method shows more significant denoising effects compared to wavelet packet decomposition and EMD methods. Additionally, the model achieved 98.7% accuracy on the preprocessed anomaly audio dataset and 99.0% on the ESC-50 dataset.

This paper innovatively proposes and combines the AMTCF-VMD preprocessing method with the Agent-SwinPyramidNet model, addressing noise interference and low accuracy issues in pipeline anomaly detection, and providing strong support for oil and gas pipeline anomaly recognition tasks in high-noise environments.