当前位置:主页 > 科技论文 > 化学工程论文 >

包含约束及噪声的相关积分方法研究及应用

发布时间:2018-03-20 10:06

  本文选题:相关积分 切入点:带约束优化 出处:《北京化工大学》2015年硕士论文 论文类型:学位论文


【摘要】:现如今,在生产工艺不断改进的同时,对于整个过程控制的水平要求也日益提高。对于工业过程的操作优化,可以不改变原有的工业装置,仅仅通过改变其操作条件实现对装置的优化,进而可以达到提高设备利用率、提高产量、增加效益的效果。一般的操作调优方法包括数学建模法和在线搜索法,但是一般这两种方法都有机理复杂建模困难、抗干扰性较差等问题。相关积分优化方法把目标函数及其调优变量都看作动力学系统进行处理,并把调优变量设置为均值可控的随机变量,因此无需建模便可实现对工业过程的操作调优且具有较强的抗干扰性。因此,本文对相关积分优化方法做了深入的研究,具体的研究内容如下所述:首先,本文研究了相关积分方法的基本原理及其递推优化过程,以及相关积分优化方法的优点及其适用条件。在原来的相关积分方法的基础上,提出了一种新的罚函数方法将输入输出约束加入到相关积分方法中去,实现在满足约束输入输出范围的条件下对工业过程操作调优,并对这种双曲罚函数的收敛性和可行性做了证明,为下文对工业过程的操作调优奠定了基础。其次本文将加约束的相关积分优化方法运用到碳二加氢反应器中。先介绍了碳二加氢反应器的反应流程,一般的碳二加氢反应器的操作调优方法并用相关积分优化方法实现了对碳二加氢反应器的操作调优,很好的解决了中石化某烯烃厂碳二加氢反应器存在的过加氢现象,提高了产品质量。最后,介绍了噪声种类对相关积分优化方法的影响,先是介绍了相关积分方法的强抗干扰能力,然后将基于核主元分析的相关积分方法运用到包含有色噪声系统的操作调优中去,能够保证系统梯度收敛到真实值附近,并可实现对系统的操作调优。
[Abstract]:Nowadays, while the production process is constantly improving, the level of control of the whole process is also increasingly required. For the optimization of the operation of the industrial process, it is possible not to change the original industrial plant. Only by changing its operating conditions, the device can be optimized, and the efficiency of equipment utilization, output and efficiency can be increased. The general operation optimization methods include mathematical modeling and online search. However, in general, these two methods have some problems such as complex modeling mechanism, poor anti-interference and so on. The objective function and its tuning variables are treated as dynamic systems by the correlation integral optimization method. The tuning variable is set as the random variable with controllable mean value, so the optimization of the industrial process can be realized without modeling and has strong anti-interference. Therefore, the related integral optimization method is deeply studied in this paper. The specific research contents are as follows: firstly, this paper studies the basic principle and recursive optimization process of the correlation integral method, as well as the advantages and applicable conditions of the correlation integral optimization method. In this paper, a new penalty function method is proposed, which adds input and output constraints to the relevant integral method, and optimizes the operation of industrial processes under the condition that the input and output range of constraints is satisfied. The convergence and feasibility of the hyperbolic penalty function are proved. It lays a foundation for the following operation optimization of industrial process. Secondly, this paper applies the constrained integral optimization method to the carbon 2 hydrogenation reactor. Firstly, the reaction process of the carbon 2 hydrogenation reactor is introduced. The general operation optimization method of carbon 2 hydrogenation reactor and the correlation integral optimization method are used to optimize the operation of carbon 2 hydrogenation reactor, which solves the phenomenon of over hydrogenation of carbon 2 hydrogenation reactor in a certain olefins plant of Sinopec. Finally, the influence of noise type on the correlation integral optimization method is introduced, first of all, the strong anti-interference ability of the correlation integral method is introduced. Then the correlation integral method based on kernel principal component analysis is applied to the operation optimization of the system with colored noise, which can ensure the gradient of the system converge to the real value and realize the operational tuning of the system.
【学位授予单位】:北京化工大学
【学位级别】:硕士
【学位授予年份】:2015
【分类号】:TQ018

【参考文献】

相关期刊论文 前5条

1 刘芳;单锐;王芳;陈延德;;双曲正弦罚函数法[J];甘肃联合大学学报(自然科学版);2008年01期

2 李绍军;张小广;;基于Alopex的进化算法及其在乙炔加氢反应器建模中的应用[J];化工自动化及仪表;2008年03期

3 张韶煜;邱代钦;穆进勇;;相关积分法优化控制技术在芳烃装置反应器的应用[J];石油化工自动化;2010年05期

4 程桂香;陈兰平;;双曲罚函数乘子法[J];首都师范大学学报(自然科学版);2007年05期

5 叶峰,邵之江,梁昔明,钱积新;Lagrange乘子初始值和罚因子迭代方式的研究[J];厦门大学学报(自然科学版);2001年S1期



本文编号:1638605

资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/huaxuehuagong/1638605.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户976ba***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com