CASIA OpenIR  > 毕业生  > 博士学位论文
流程工业MES中若干问题的研究
其他题名Research on several MES problems in Process Industry
邓全亮
2006-05-30
学位类型工学博士
中文摘要随着企业信息化要求越来越高,企业的综合信息集成是发展的必然趋势。底层的控制系统发展得很成熟了,并已广泛应用到各行各业,为实现生产的自动化打下了良好的基础;上层的管理软件ERP(企业资源计划)也应用得越来越广泛,成为企业上层管理的必然要求。企业要灵活生产,下层的控制系统必须能得到上层及时的指导信息;企业要迅速决策,也必须能及时得到下层的生产反馈信息;但是,目前企业上层管理系统与下层的控制系统之间的信息流的不通畅,已经成为了信息交流的瓶颈,极大地影响了企业综合信息化的效果。而MES(制造执行系统)正是为解决这个问题而提出来的。 本论文以国家计委“基于多现场总线及Internet的综合自动化系统MFIAS”项目及国家“十一五”项目――石化MES的研究子课题为背景,对流程工业MES中的若干问题进行了研究。 本文主要内容和创新之处如下: 1. 数据集成模型的研究及应用。分析了流程工业发展特点,针对底层系统纷繁复杂的情况,考虑到对遗留系统的保护和整个系统的开放性,提出了基于中间件(ODBC中间件和基于OPC技术的中间件)的数据采集模型。并基于此模型在多个工厂实施了系统集成项目,并取得了良好的效果。 2. 流程工业MES中优化模块的研究。根据焦化企业配煤炼焦的具体特点,对配煤优化进行了研究,利用统计学方法和神经网络建模方法建立配合煤和焦炭质量预测模型,利用遗传算法对配煤比进行优化,提出了基于预测模型的配煤优化方法,并通过仿真实验验证此方法的可行性。 3. 配煤调度的实现。提出了在炼焦配煤中从调度令下达,到配煤计划的优化及优化策略在控制系统实施的方法,并给出了整个过程的程序设计框架。 4. MES系统集成框架的研究。分析了流程工业特点及其对MES系统的要求,结合多Agent技术和CORBA的特点,提出了基于MAS系统集成框架和基于CORBA的系统通信机制,并通过一个实例分析其工作流程。
英文摘要With the development of informationization, MES has become an inevitable trend in enterprise information integration. PCS (Process Control System),which has been a mature product and has been widely applied to all kinds of industry, lays a solid foundation for production automation. Management software like ERP is being used in growing numbers in top layer of many enterprises. To gain production agility, the control systems in the bottom layer need timely information from top level. On the other hand, Managers on top level need timely feedbacks from bottom layer for effective decision-making. Now the information bottleneck between the PCS layer and Management layer is exerting a negative effect on enterprise informationization of enterprise. MES is intended to solve this problem. This dissertation is based on the project “Multi-fieldbus and Internet Based Integrated Automation System”of integrated automation center and the sub subject of petrochemical MES of national “eleven five” planning. The major contents and innovation are as follows: 1. Research and application of data integration model. A middleware based data integration model is put forward, which takes full consideration of the characteristics of process industry, the various control system on bottom layer, the reservation of old system and the opening of the system. This model was validated in practical project and brought good results. 2. Research on optimization. Research was made on coal blending optimization according to the characteristics of coal blending and coking. Optimization model based on predictive model on coal blending and coking is developed, which is proved to be effective through simulation. 3. Implementation of coal blending scheduling. A detailed method is given on how to realize coal blending scheduling. Which covers the transferring of instruction, the optimization method and how this method is implemented in control system. Also the whole programming framework is given. 4. Research of MES integration framework. According to the characteristic of process industry and the requirement of MES, a MES integration framework based on MAS and CORBA is put forward. The workflow of the framework is shown through a case.
关键词Mes 系统集成 数据集成 Opc Odbc 优化 遗传算法 神经网络 Mas Corba Mes System Integration Data Integration Opc Odbc Optimization Genetic Algorithm Neural Network Mas Corba
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/5921
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
邓全亮. 流程工业MES中若干问题的研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2006.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
CASIA_20031801460296(2063KB) 暂不开放CC BY-NC-SA
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[邓全亮]的文章
百度学术
百度学术中相似的文章
[邓全亮]的文章
必应学术
必应学术中相似的文章
[邓全亮]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。