A general numerical method for solving the three-dimensional hyperbolic heat conduction equation on unstructured grids

He, Huizhi; Zhang, Xiaobing

doi:10.1016/j.camwa.2024.01.012

Cited by 3 publications

(1 citation statement)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To date, the optimization research on cloud service load balancing has not fully explored the possibility and potential advantages of integrating physical models, such as the heat conduction equation, with machine learning algorithms [9,10].…”

Section: Introductionmentioning

confidence: 99%

Optimizing Cloud Service Load Balancing Through Heat Conduction Equation Applications

He,

Wang,

Liu

et al. 2024

IJHT

View full text Add to dashboard Cite

With the rapid development of cloud computing, load balancing technology in cloud services has become a critical component in ensuring service quality and system stability. Traditional load balancing methods, often relying on static parameters and preset rules, face challenges in flexibly responding to the dynamic changes in cloud service demands. Recent studies have begun to explore the application of machine learning algorithms to optimize load balancing, aiming to enhance the system's adaptive adjustment capabilities. This study proposes an innovative approach, applying the solution of the heat conduction equation to the optimization of cloud service load balancing issues. It simulates and analyzes the dynamic changes in load distribution, proposing corresponding optimization strategies. The first part of this research focuses on designing a model that integrates genetic algorithms and neural networks to solve the inverse problem of the twodimensional nonlinear heat conduction equation, namely, the accurate prediction of thermal physical parameters. By simulating the heat conduction process, this model can reflect the dynamic distribution characteristics of server loads and guide the adjustment strategy of weights. Furthermore, an adaptive dynamic load balancing strategy algorithm is proposed. By optimizing the existing engine x (Nginx) weighted least connections algorithm, an efficient adaptive algorithm is designed and implemented. This algorithm adjusts server weights dynamically based on real-time load data, enabling cloud services to respond more flexibly and efficiently to different service requests. The findings of this research not only enhance the processing capability and resource utilization rate of cloud services but also provide more scientific and precise theoretical support for load balancing through the introduction of new algorithmic models. Additionally, the proposed adaptive dynamic load balancing strategy algorithm has demonstrated good performance in practical deployment, offering new perspectives and technical paths for the research and practice of cloud service load balancing.

show abstract

Section: Introductionmentioning

confidence: 99%