Cloud computing is a thriving and ever-expanding sector in the industry world. This growth has sparked increased interest from organizations seeking to harness its potential. However, the sheer volume of services and offerings in this field has resulted in a noticeable surge in related data. With the rapid evolution and growing demand, cloud computing resource management faces a fresh set of challenges. Resource limitations, such as high maintenance costs, elevated Energy Consumption (EC), and adherence to Service Level Agreements (SLA), are critical concerns for both the cloud computing industry and its user organizations. In this context, taking a proactive approach to resource management and Virtual Machine Consolidation (VMC) has become imperative. The logical management of resources and the consolidation of Virtual Machines (VMs) in a manner that aligns with the requirements and demands of service providers and users have garnered widespread attention. The goal of this proposed paper is to focus on addressing the VMC problem within a unified framework, divided into two main phases. The first phase deals with host workload detection and prediction, while the subsequent phase tackles the selection and allocation of appropriate VMs. In our proposed method, for the first time, we use a Granular Computing (GRC) model, which is an efficient, scalable, and human-centric computational approach. This model exhibits behaviors similar to intelligent human decision-making, as it can simultaneously consider all factors and criteria involved in the problems. We evaluated our proposed method through simulations using CloudSim on various types of workloads. Experimental results demonstrate that our proposed algorithm outperforms other algorithms in all measurement metrics.

中文翻译：

---

高效的云数据中心：动态虚拟机整合的自适应框架

云计算是工业界一个蓬勃发展且不断扩展的领域。这种增长引发了寻求利用其潜力的组织越来越大的兴趣。然而，该领域的服务和产品数量庞大，导致相关数据显着激增。随着快速发展和不断增长的需求，云计算资源管理面临着一系列新的挑战。资源限制，例如高维护成本、增加的能源消耗 (EC) 以及遵守服务级别协议 (SLA)，是云计算行业及其用户组织的关键问题。在这种背景下，采取主动的资源管理和虚拟机整合 (VMC) 方法已势在必行。以符合服务提供商和用户需求的方式对资源进行逻辑管理以及对虚拟机 (VM) 进行整合已引起广泛关注。本文的目标是集中解决统一框架内的 VMC 问题，分为两个主要阶段。第一阶段处理主机工作负载检测和预测，而后续阶段处理适当虚拟机的选择和分配。在我们提出的方法中，我们首次使用粒度计算（GRC）模型，这是一种高效、可扩展且以人为中心的计算方法。该模型表现出类似于人类智能决策的行为，因为它可以同时考虑问题中涉及的所有因素和标准。我们通过使用 CloudSim 对各种类型的工作负载进行模拟来评估我们提出的方法。实验结果表明，我们提出的算法在所有测量指标上都优于其他算法。