【摘要】 嵌入式操作系统虚拟化技术因其标准化的开发流程和对跨平台的适应能力受到了广泛关注。Docker作为一种轻量级虚拟化方式，具有资源利用率高，启动速度快，实时性好的特点，是嵌入式操作系统高效的虚拟化方案。文章介绍了Docker的原理，并与传统虚拟化技术进行了对比；以智能电能表为例提出了其功能容器化的方案，在Docker容器中构造了基于二叉树支持向量机(BT-SVM)的电能质量扰动识别分类器，实现了7种电能质量扰动的识别，并且测试了其相对传统虚拟机的轻量级特性。为智能量测设备嵌入式操作系统虚拟化与其功能部署容器化提供了一种思路。更多还原

【Abstract】 The embedded operating system virtualization technology has attracted extensive attention because of its standardized development process and adaptability of cross-platform. As a lightweight virtualization method, Docker has the characteristics of high resource utilization, fast startup speed and good real-time performance. It is an efficient virtualization scheme for embedded electric energy meter operating system. This paper introduces the principle of Docker and compares Docker with traditional virtualization technology. Puts forward the scheme of power meter function containerization and construct a power quality disturbance recognition classifier based on Binary Tree Support Vector Machine(BT-SVM), realize the identification of seven power quality disturbances in Docker environment. And test its lightweight characteristics compared with traditional virtual machine. It provides an idea for the virtualization of smart measurement equipment operating system and the containerization of function deployment of intelligent measuring equipment.更多还原