结构健康监测物联网系统的云计算应用研究

发布时间：2018-04-05 14:17

本文选题：结构健康监测　切入点：物联网　出处：《哈尔滨工业大学》2013年硕士论文

【摘要】：结构健康监测系统是保障重大工程结构安全的重要手段，对于保障城市安全和城市功能的正常发挥以及经济和社会的可持续发展具有重大意义。实现结构健康监测与物联网的融合是数字城市建设的必然要求，可促进结构健康监测发展，使其发挥更大作用。云计算技术是物联网海量感知数据高效智能处理的重要支撑技术，将物联网和云计算技术应用到结构健康监测领域是一个重要的研究方向。为了实现物联网在结构健康监测领域的应用，本文根据结构健康监测的要求，提出了结构健康监测物联网的总体方案；考虑云计算对该系统的支持，分析了结构健康监测中加速度响应信号处理等任务的可并行性，利用分析结果对结构健康监测云中的计算与存储进行了概念设计；为了验证云计算的应用分析，利用MATLAB对ITD模态参数识别任务进行了分布式计算模拟。本文具体完成了以下三方面工作：提出了结构健康监测物联网系统的总体方案，并基于功能需求分析了各功能子层的概念设计，，同时设计了C/S与B/S混合的结构健康监测物联网管理系统体系结构。分析并得到了结构健康监测基于加速度响应的模态参数识别、损伤诊断、模型修正和安全评定等任务的可并行性分析结果，也分析了相关任务的分解、计算与存储策略以及云中分布式计算适用性。并利用上述分析结果对结构健康监测云中的计算与存储进行了概念设计，主要考虑云中计算与存储的整合和迁移；针对纵向数据分块策略，举例部署了ITD模态参数识别和有限元模型修正任务在云中计算与存储的整合和迁移及其执行控制。利用MATLAB并行和分布式计算工具箱对ITD模态参数识别任务进行了云计算模拟，模拟实现了pmode、parfor、SPMD共3种并行结构5个并行计算程序的单机和多机云中分布式计算。并将其与单机串行计算对比，分析了模拟分布式计算对计算效率的提高，结果显示其中的通信代价不可忽略，同时对模拟分布式计算与云中分布式计算进行了理论对比分析，着重分析了两者的不同，通过对比得出了云中分布式计算更高效更稳定可靠的结论。
[Abstract]:Structural health monitoring system is an important means to ensure the safety of major engineering structures. It is of great significance to ensure the safety of cities, the normal exertion of urban functions and the sustainable development of economy and society.The integration of structural health monitoring and Internet of things is an inevitable requirement of digital city construction, which can promote the development of structural health monitoring and make it play a greater role.Cloud computing technology is an important supporting technology for efficient and intelligent processing of mass perceptual data in the Internet of things. It is an important research direction to apply the technology of Internet of things and cloud computing to the field of structural health monitoring.In order to realize the application of Internet of things in the field of structural health monitoring, according to the requirements of structural health monitoring, this paper puts forward the overall scheme of structural health monitoring of Internet of things, considering the support of cloud computing to the system.The parallelism of acceleration response signal processing in structural health monitoring is analyzed. The calculation and storage in structural health monitoring cloud are designed by using the analysis results.MATLAB is used to simulate the ITD modal parameter identification task.This paper has completed the following three aspects of work:In this paper, the overall scheme of structured health monitoring and the concept design of each functional sub-layer are analyzed based on the functional requirements, and the architecture of the structured health monitoring system of IoT based on the mixture of C / S and B / S is designed at the same time.The parallelism analysis results of modal parameter identification, damage diagnosis, model modification and safety assessment of structural health monitoring based on acceleration response are analyzed and the decomposition of related tasks is also analyzed.Computing and storage strategy and applicability of distributed computing in the cloud.The conceptual design of calculation and storage in the structural health monitoring cloud is carried out by using the above analysis results, mainly considering the integration and migration of the calculation and storage in the cloud, and aiming at the strategy of dividing the longitudinal data into blocks,For example, the integration and migration of ITD modal parameter identification and finite element model modification in cloud computing and storage and its execution control are deployed.The MATLAB parallel and distributed computing toolbox is used to simulate the ITD modal parameter identification task. The simulation realizes the distributed computing in single computer and multi-machine cloud with three parallel structures and five parallel computing programs.By comparing it with single computer, the paper analyzes the improvement of computing efficiency by simulating distributed computing. The results show that the communication cost can not be ignored. At the same time, the theoretical comparison between simulated distributed computing and cloud distributed computing is carried out.The difference between them is analyzed, and the conclusion that distributed computing in cloud is more efficient, more stable and reliable is obtained by comparison.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.44;TN929.5;TU317

【相似文献】