基于虚拟化容器和AI算法引擎的办公云系统研究  

Research on Office Cloud System Based on Virtualization Container and AI Algorithm Engine

作　　者：农英雄[1,2] 陈智斌[1] 梁冬[1] 李喆 汪倍贝[1] NONG Yingxiong;CHEN Zhibin;LIANG Dong;LI Zhe;WANG Beibei(China Tobacco Guangxi Industrial Co.,Ltd.,Nanning 530001,China;Guangxi University,Nanning 530004,China)

机构地区：[1]广西中烟工业有限责任公司,南宁530001 [2]广西大学,南宁530004

出　　处：《自动化与仪器仪表》2024年第11期194-197,共4页Automation & Instrumentation

基　　金：国家重点研发计划项目《私有云环境下服务化智能办公系统平台》(2018YFB1004101)。

摘　　要：为在实现现代化办公的同时进一步提升办公效率和质量,以烟草企业办公为例,提出一种基于虚拟化容器和人工智能算法的办公云系统。其中,基于Docker虚拟化技术进行办公云系统的构建,同时引入卷积神经网络以及长短时记忆网络进行资源预测模型的构建,并对模型进行对应的改进,以提升资源预测的精度,进而提升办公系统的管理水平。实验结果表明,与其他类型的预测模型相比,本研究所构建的基于改进LSTM融合CNN模型的资源预测模型具有更高的预测性能,在RMSE和MAE两个评价指标上均表现出更高的精度;将所构建的办公云系统应用于实际的烟草企业办公场景中时,能够进行准确的CPU使用率预测,预测结果十分接近实际的CPU使用情况。综上,所构建的基于虚拟化容器和人工智能算法的办公云系统性能良好,将其应用于实际的烟草企业的日常办公场景中,能够帮助使用者提升数据处理水平,进而提升工作效率和工作质量,可行性较高。To further improve office efficiency and quality while achieving modern office work,taking tobacco enterprise office as an example,a virtual container and artificial intelligence algorithm based office cloud system is proposed.Among them,the construction of an office cloud system is based on Docker virtualization technology,and convolutional neural networks and long short-term memory networks are introduced to construct a resource prediction model.Corresponding improvements are made to the model to improve the accuracy of resource prediction and thus enhance the management level of the office system.The experimental results show that compared with other types of prediction models,the resource prediction model based on the improved LSTM fused CNN model constructed in this study has higher predictive performance,and shows higher accuracy in both RMSE and MAE evaluation indicators;When applying the constructed office cloud system to actual tobacco enterprise office scenarios,accurate CPU usage predictions can be made,and the predicted results are very close to the actual CPU usage.In summary,the office cloud system based on virtualization containers and artificial intelligence algorithms constructed has good performance.Applying it to the daily office scenarios of actual tobacco enterprises can help users improve their data processing level,thereby improving work efficiency and quality,and has high feasibility.

关 键 词：智能办公 虚拟化技术 DOCKER CNN LSTM 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]