Evacuation planning is important for reducing casualties in toxic gas leak incidents. However, most evacuation plans are too qualitative to be applied to unexpected practical situations. Here, we suggest an evacuation...
详细信息
Evacuation planning is important for reducing casualties in toxic gas leak incidents. However, most evacuation plans are too qualitative to be applied to unexpected practical situations. Here, we suggest an evacuation route proposal system based on a quantitative risk evaluation that provides the safest route for individual evacuees by predicting dynamic gas dispersion with a high accuracy and short calculation time. Detailed evacuation scenarios, including weather conditions, leak intensity, and evacuee information, were considered. The proposed system evaluates the quantitative risk in the affected area using a deep neural network surrogate model to determine optimal evacuation routes by integer programming. The surrogate model was trained using data from computational fluid dynamics simulations. A variational autoencoder was applied to extract the geometric features of the affected area. The predicted risk was combined with linearized integer programming to determine the optimal path in a predefined road network. A leak scenario of an ammonia gas pipeline in a petrochemical complex was used for the case study. The results show that the developed model offers the safest route within a few seconds with minimum risk. The developed model was applied to a sensitivity analysis to determine variable influences and safe shelter locations.
Automatic identification system (AIS) refers to a new type of navigation aid system equipped in maritime vehicles to monitor ship performance. It provides trajectory information of vessels which can be used for the cl...
详细信息
Automatic identification system (AIS) refers to a new type of navigation aid system equipped in maritime vehicles to monitor ship performance. It provides trajectory information of vessels which can be used for the classification task. The classification results facilitate ocean surveillance and conservation, vessel management enhancement, fishery regulation, marine ecological sustainability, and nautical safety protection. Some progresses have been made based on machine learning or deep learning strategies to perform supervised learning by assuming that during the training process, the category labels of historical trajectory data are available. However, in reality, the label information may be difficult or expensive to obtain, resulting in that only a small part of the training data is labeled, and most of the training data is unlabeled. To address this issue, this paper proposes a semi-supervised deep learning approach to integrate the knowledge of unlabeled data for vessel trajectory classification. Here, we call our approach SSL-VTC. Specifically, we first extract vessel trajectories by integrating the kinematic and static information from historical AIS messages. Then, we design convolutional neural networks to extract feature representations from the vessel trajectories. Finally, we develop a semi-supervised learning algorithm based on the variational autoencoder to perform discriminative learning and generative learning simultaneously. In this way, our SSL-VTC framework can fully leverage the labeled data and unlabeled data during the training process. To the best of our knowledge, we are the first to utilize semi-supervised learning for vessel trajectory classification. Experimental results on the public AIS dataset show that our SSL-VTC can effectively extract feature representations from the AIS messages and its performance is significantly better than the traditional supervised learning methods. The approach and findings of this study provide important i
Climate change has increased the intensity and frequency of storms in many world regions, calling for new flood planning and management strategies. The concept of flood drainage rights (FDR), or the legal rights of re...
详细信息
Climate change has increased the intensity and frequency of storms in many world regions, calling for new flood planning and management strategies. The concept of flood drainage rights (FDR), or the legal rights of regions to drain floodwaters into river reaches, is used in watershed planning in China. Quantifying the allocation of FDR remains challenging, where some previous methods have resulted in unreasonable or impractical allocation plans due to incomplete consideration of driving factors or the use of unscientific allocation methods. This study ex-plores the allocation plan of FDR in the middle and lower reaches of the Yellow River Watershed in China. Climatic variability and change have caused frequent flooding in portions of the basin, with significant societal and economic implications. First, we comprehensively analyzed factors driving FDR for regions in the watershed. Following the conceptual flood resilience strategy currently being advocated for the region, we considered natural, socioeconomic, governance, resilience, and resistance factors that influence the complex allocation of FDR and established a qualitative indicator system to reflect the complexity of these driving factors. Second, we quantified FDR values for flood-prone regions in the middle and lower river reaches of this major river basin. We introduced a specific deep learning method, called the variational autoencoder (VAE) model, to quantify FDR allocation, providing a robust solution to the challenge of the multi-objective, high-dimensional, nonlinear, and non-normal distribution of factors driving FDR allocation. Next, using data from 2005 to 2019, this model was applied to the study area. The allocation of FDR (summing to 100%) across five flood-prone provinces of the watershed includes Inner Mongolia (9.36%), Shaanxi (10.00%), Shanxi (10.95%), Henan (32.58%), and Shan-dong (37.12%). Using the harmony evaluation method based on harmony theory, we compared the new VAE allocation method
Machine Learning (ML) techniques have been used in an extensive range of applications in the field of structural and multidisciplinary optimization over the last few years. This paper presents a survey of this wide bu...
详细信息
Machine Learning (ML) techniques have been used in an extensive range of applications in the field of structural and multidisciplinary optimization over the last few years. This paper presents a survey of this wide but disjointed literature on ML techniques in the structural and multidisciplinary optimization field. First, we discuss the challenges associated with conventional optimization and how Machine learning can address them. Then, we review the literature in the context of how ML can accelerate design synthesis and optimization. Some real-life engineering applications in structural design, material design, fluid mechanics, aerodynamics, heat transfer, and multidisciplinary design are summarized, and a brief list of widely used open-source codes as well as commercial packages are provided. Finally, the survey culminates with some concluding remarks and future research suggestions. For the sake of completeness, categories of ML problems, algorithms, and paradigms are presented in the Appendix.
Esophageal disorders are related to the mechanical properties and function of the esophageal wall. Therefore, to understand the underlying fundamental mechanisms behind various esophageal disorders, it is crucial to m...
详细信息
Esophageal disorders are related to the mechanical properties and function of the esophageal wall. Therefore, to understand the underlying fundamental mechanisms behind various esophageal disorders, it is crucial to map mechanical behavior of the esophageal wall in terms of mechanics-based parameters corresponding to altered bolus transit and increased intrabolus pressure. We present a hybrid framework that combines fluid mechanics and machine learning to identify the underlying physics of various esophageal disorders (motility disorders, eosinophilic esophagitis, reflux disease, scleroderma esophagus) and maps them onto a parameter space which we call the virtual disease landscape (VDL). A one-dimensional inverse model processes the output from an esophageal diagnostic device called the functional lumen imaging probe (FLIP) to estimate the mechanical "health" of the esophagus by predicting a set of mechanics-based parameters such as esophageal wall stiffness, muscle contraction pattern and active relaxation of esophageal wall. The mechanics-based parameters were then used to train a neural network that consists of a variational autoencoder that generated a latent space and a side network that predicted mechanical work metrics for estimating esophagogastric junction motility. The latent vectors along with a set of discrete mechanics-based parameters define the VDL and formed clusters corre-sponding to specific esophageal disorders. The VDL not only distinguishes among disorders but also displayed disease progression over time. Finally, we demonstrated the clinical applicability of this framework for esti-mating the effectiveness of a treatment and tracking patients' condition after a treatment.
Topology optimization is a powerful methodology for generating novel designs with a high degree of design freedom. In exchange for this attractive feature, topology optimization cannot generally avoid multimodality, w...
详细信息
Topology optimization is a powerful methodology for generating novel designs with a high degree of design freedom. In exchange for this attractive feature, topology optimization cannot generally avoid multimodality, which often impedes finding a satisfactory solution when dealing with strongly nonlinear optimization problems. In this study, we focus on constructing a framework that aims to indirectly solve such complex topology optimization problems. The framework is based on multifidelity topology design (MFTD), the basic concept of which is to divide solving an original topology optimization problem into two kinds of procedures, i.e., low-fidelity optimization and high-fidelity evaluation. We build the framework as a data-driven approach, where the design candidates given by solving the low-fidelity optimization problem are iteratively updated based on the idea of the evolutionary algorithm (EA). As a key procedure to realize a crossover-like operation in the high-dimensional design space, a variational autoencoder-one of the representative deep generative models-is utilized for generating a new dataset composed of various material distributions. Besides, we propose a mutation-like operation for generating novel material distributions based on reference ones in the dataset. By integrating these operations with an elitism-based selection procedure, we propose data-driven MFTD that enables gradient-free optimization even if tackling a complex optimization problem with a high degree of design freedom. We apply the proposed framework to forced convection heat transfer problems, where the low-fidelity optimization problem is formulated using a Darcy flow model, whereas the high-fidelity evaluation is conducted using a Navier-Stokes model. We first demonstrate that the obtained results from the proposed framework can achieve almost identical performance compared with that of the directly solved results in a well-known 2D laminar heat transfer problem. We then show that
This work presents a methodology for the generation of novel 3D objects resembling wireframes of building types. These result from the reconstruction of interpolated locations within the learnt distribution of variati...
详细信息
This work presents a methodology for the generation of novel 3D objects resembling wireframes of building types. These result from the reconstruction of interpolated locations within the learnt distribution of variational autoencoders (VAEs), a deep generative machine learning model based on neural networks. The data set used features a scheme for geometry representation based on a 'connectivity map' that is especially suited to express the wireframe objects that compose it. Additionally, the input samples are generated through 'parametric augmentation', a strategy proposed in this study that creates coherent variations among data by enabling a set of parameters to alter representative features on a given building type. In the experiments that are described in this paper, more than 150 k input samples belonging to two building types have been processed during the training of a VAE model. The main contribution of this paper has been to explore parametric augmentation for the generation of large data sets of 3D geometries, showcasing its problems and limitations in the context of neural networks and VAEs. Results show that the generation of interpolated hybrid geometries is a challenging task. Despite the difficulty of the endeavour, promising advances are presented.
The aim of this paper was the detection of pathologies through respiratory sounds. The ICBHI (International Conference on Biomedical and Health Informatics) Benchmark was used. This dataset is composed of 920 sounds o...
详细信息
The aim of this paper was the detection of pathologies through respiratory sounds. The ICBHI (International Conference on Biomedical and Health Informatics) Benchmark was used. This dataset is composed of 920 sounds of which 810 are of chronic diseases, 75 of non-chronic diseases and only 35 of healthy individuals. As more than 88% of the samples of the dataset are from the same class (Chronic), the use of a variational Convolutional autoencoder was proposed to generate new labeled data and other well known oversampling techniques after determining that the dataset classes are unbalanced. Once the preprocessing step was carried out, a Convolutional Neural Network (CNN) was used to classify the respiratory sounds into healthy, chronic, and non-chronic disease. In addition, we carried out a more challenging classification trying to distinguish between the different types of pathologies or healthy: URTI, COPD, Bronchiectasis, Pneumonia, and Bronchiolitis. We achieved results up to 0.993 F-Score in the three-label classification and 0.990 F-Score in the more challenging six-class classification.
Recently, adversarial examples become one of the most dangerous risks in deep learning, which affects applications of real world such as robotics, cyber-security and computer vision. In image classification, adversari...
详细信息
ISBN:
(纸本)9781510640412
Recently, adversarial examples become one of the most dangerous risks in deep learning, which affects applications of real world such as robotics, cyber-security and computer vision. In image classification, adversarial attacks showed the ability to fool classifiers with small imperceptible perturbations added to the input. In this paper, we present an efficient defense mechanism, we call DVAE-SR that combine variational autoencoder and super-resolution to eliminate adversarial perturbation from image input before feeding it to the CNN classifier. The DVAE-SR can successfully defend against both white-box and black-box attacks without retraining CNN classifier and it recovers better accuracy than Defense-GAN and Defense-VAE..
This article presents the process of building a system generating music content of a specified emotion. As the emotion labels, four basic emotions: happy, angry, sad, relaxed, which correspond to the four quarters of ...
详细信息
ISBN:
(纸本)9781665442077
This article presents the process of building a system generating music content of a specified emotion. As the emotion labels, four basic emotions: happy, angry, sad, relaxed, which correspond to the four quarters of Russell's model, were used. Conditional variational autoencoder using a recurrent neural network for sequence processing was used as a generative model. The obtained results in the form of the generated music examples with a specific emotion are convincing in their structure and sound. The generated examples were evaluated through comparison with the training set.
暂无评论