Depleting coal reserves and increasing carbon emissions have propelled the use of renewable energy sources to generate electricity. Microgrids have been recognised as one of the solutions to provide continuity of powe...
详细信息
The grid resolution would affect the pollution concentrations and thus on the Air Quality Indices (AQI) - a generalized assessment of the air quality impact on human health. Therefore, we made a numerical experiment f...
详细信息
ISBN:
(纸本)9783031562075;9783031562082
The grid resolution would affect the pollution concentrations and thus on the Air Quality Indices (AQI) - a generalized assessment of the air quality impact on human health. Therefore, we made a numerical experiment for evaluation of the horizontal grid resolution impact on the simulated AQI over the territory of Bulgaria. We used a set of models used worldwide - WRF - the meteorological preprocessor, CMAQ - chemical transport model, SMOKE - emission model for performing computer simulations. The NCEP Global Analysis Data with a horizontal resolution of 1 degrees x 1 degrees are used as a background meteorological data used in the study. Using the "nesting" capabilities of the WRF and CMAQ models, a resolution of 9 km was achieved for the territory of Bulgaria, by sequentially solving the task in several consecutive, nested areas. The simulations were performed for three cases for grid and emission resolutions for the period 2008-2014, creating an ensemble large and comprehensive enough, reflects the most typical atmospheric conditions with their typical recurrence. The spatial/temporal distribution of the recurrence of the different AQI categories for Bulgaria are calculated. Comparing them for the above 3 cases makes it possible to evaluate the grid resolution impact.
In healthcare, safeguarding patient data integrity is paramount, necessitating robust deep learning models. This research focuses on bolstering the cyber security defenses of these models, particularly against adversa...
详细信息
ISBN:
(纸本)9798350372816
In healthcare, safeguarding patient data integrity is paramount, necessitating robust deep learning models. This research focuses on bolstering the cyber security defenses of these models, particularly against adversarial attacks, through the application of the Localized Adversarial Feature Attack (LAFEAT) technique. Leveraging Graphics Processing Units (GPU) and the TensorFlow framework, the study explores algorithmic enhancements like the Fast Gradient Sign Method (FGSM) and Model-Agnostic Meta-Learning (MAML). These techniques perturb input data gradients or adapt model parameters, enhancing resilience against adversarial perturbations. Additionally, Particle Swarm Optimization (PSO), a metaheuristic optimization technique, is examined to fortify defense mechanisms further. parallel processing techniques, utilizing parallel GPU or distributed clusters, are implemented to expedite the optimization process, reducing computational burdens and enhancing scalability. Despite advancements, gaps remain in understanding how these models can be optimized for real-world healthcare applications, particularly in terms of balancing computational efficiency and robustness. The research demonstrates tangible benefits, with a significant 10% increase in success rates against adversarial attacks compared to baseline methods. Moreover, parallel processing implementation results in a 30% reduction in optimization time, improving the efficiency of cyber security defenses for deep learning models in healthcare. These numerical enhancements underscore the research's value in fortifying healthcare systems against adversarial threats, ensuring the security and integrity of patient data while addressing existing gaps in the optimization of adversarial defenses in healthcare applications. The strengths of this research lie in its robust adversarial attack optimization techniques, integration of parallel processing for efficiency, and application to critical healthcare tasks. Comprehensive
We introduce SpDISTAL, a compiler for sparse tensor algebra that targets distributed systems. SpDISTAL combines separate descriptions of tensor algebra expressions, sparse data structures, data distribution, and compu...
详细信息
As more and more services in the power grid introduce machine learning and deep learning technologies, feature engineering has become more and more complex. In order to improve the efficiency of feature engineering, t...
详细信息
Modern day power systems have contributed to climatic change and reduced efficiency as a result of more utilization of conventional energy resources to match the increasing power demand. Hence grid systems that use di...
详细信息
The manual deployment of applications distributed across the cloud, fog, and edge is error-prone and complex. TOSCA is a standard for modeling the deployment of cloud applications in a vendor-neutral and technology-in...
详细信息
ISBN:
(纸本)9798400702341
The manual deployment of applications distributed across the cloud, fog, and edge is error-prone and complex. TOSCA is a standard for modeling the deployment of cloud applications in a vendor-neutral and technology-independent manner that is also suitable for the fog and edge continuum. However, there exist various TOSCA orchestrators with different functionalities. Thus, selecting an appropriate TOSCA orchestrator requires technical expertise since all the available orchestrators must be analyzed regarding technical, functional, legal, and organizational requirements. In this paper, we tackle this issue and present a systematic technology review of TOSCA orchestrators. Our goal is to support project managers, developers, and researchers in selecting a suitable TOSCA orchestrator. For this, we select actively maintained general-purpose open-source TOSCA orchestrators. Moreover, we introduce the TOSCA Orchestrator Classification Framework and present a selection support system.
Major equipment manufacturers are increasing their research and development investment, and have successively launched solutions for 100G (Gigabyte) fiber optic transmission systems. However, the 100G DP-QPSK (Dual-po...
详细信息
In the increasingly advanced environment of technology, the power distribution network is gradually developing towards automation, and the functions of its distribution automation terminal equipment are also becoming ...
详细信息
When solving compute-intensive tasks, CPU/GPU hardware resources and specialized grid, Custer, Cloud infrastructure are commonly used to achieve high performance. However, this requires a high initial capital expense ...
详细信息
ISBN:
(纸本)9783031061561;9783031061554
When solving compute-intensive tasks, CPU/GPU hardware resources and specialized grid, Custer, Cloud infrastructure are commonly used to achieve high performance. However, this requires a high initial capital expense and ongoing maintenance costs. In contrast, ARM-based mobile devices regularly see improvement in their capacity, stability, and processing power daily while becoming ever more ubiquitous and requiring no massive capital or operating expenditures thanks to their reduced size and energy efficiency. Given this shifting computer paradigm, it is conceivable that a cost- and power-efficient solution for our world's HPC processing tasks would include ARM-based mobile devices, while they are idle during recharging periods. We proposed, developed, deployed and evaluated a distributed, collaborative, elastic and low-cost platform to solve HPC tasks recycling ARM mobile resources based on Cloud, microservices and containers, efficiently orchestrated via Kubernetes. To validate the system scalability, flexibility, and performance a lot of concurrent video transcoding scenarios were run. The results showed the system allows for improvements in terms of scalability, flexibility, stability, efficiency, and cost for HPC workloads.
暂无评论