计算机与应用化学
計算機與應用化學
계산궤여응용화학
COMPUTERS AND APPLIED CHEMISTRY
2013年
6期
629-632
,共4页
谭超%王超%李惟一%吴同%张开仕
譚超%王超%李惟一%吳同%張開仕
담초%왕초%리유일%오동%장개사
精馏%推断控制%仿真%MATLAB
精餾%推斷控製%倣真%MATLAB
정류%추단공제%방진%MATLAB
distillation%inferential control%simulation%MATLAB
在化工生产实践中,基于被控变量检测的反馈控制系统占有绝对比重。不过,实际生产过程中却存在着这样一类情况,过程的被控制变量,甚至过程的扰动均无法测量或难以测量,如精馏塔塔顶、塔底产品的组成,因而难于实现反馈控制或前馈控制。精馏塔是化工等行业中广泛使用的分离设备,其控制方案在化工过程中具有十分关键的作用。为满足工艺要求和节能,需将塔顶和塔底产品流控制在设计值。理论上,将产品组成直接作为被控变量是最佳的,但仍有一些问题限制了其在实践中的推广使用。基于可测辅助变量推断难以直接测量或测量滞后太大的关键过程变量的推断控制能弥补这一缺陷。本文以多组分精馏过程为例,在 MATLAB 平台进行了多变量推断控制控制设计及仿真,并对其鲁棒性进行了分析,达到较好的控制仿真效果。
在化工生產實踐中,基于被控變量檢測的反饋控製繫統佔有絕對比重。不過,實際生產過程中卻存在著這樣一類情況,過程的被控製變量,甚至過程的擾動均無法測量或難以測量,如精餾塔塔頂、塔底產品的組成,因而難于實現反饋控製或前饋控製。精餾塔是化工等行業中廣汎使用的分離設備,其控製方案在化工過程中具有十分關鍵的作用。為滿足工藝要求和節能,需將塔頂和塔底產品流控製在設計值。理論上,將產品組成直接作為被控變量是最佳的,但仍有一些問題限製瞭其在實踐中的推廣使用。基于可測輔助變量推斷難以直接測量或測量滯後太大的關鍵過程變量的推斷控製能瀰補這一缺陷。本文以多組分精餾過程為例,在 MATLAB 平檯進行瞭多變量推斷控製控製設計及倣真,併對其魯棒性進行瞭分析,達到較好的控製倣真效果。
재화공생산실천중,기우피공변량검측적반궤공제계통점유절대비중。불과,실제생산과정중각존재착저양일류정황,과정적피공제변량,심지과정적우동균무법측량혹난이측량,여정류탑탑정、탑저산품적조성,인이난우실현반궤공제혹전궤공제。정류탑시화공등행업중엄범사용적분리설비,기공제방안재화공과정중구유십분관건적작용。위만족공예요구화절능,수장탑정화탑저산품류공제재설계치。이론상,장산품조성직접작위피공변량시최가적,단잉유일사문제한제료기재실천중적추엄사용。기우가측보조변량추단난이직접측량혹측량체후태대적관건과정변량적추단공제능미보저일결함。본문이다조분정류과정위례,재 MATLAB 평태진행료다변량추단공제공제설계급방진,병대기로봉성진행료분석,체도교호적공제방진효과。
In the process of chemical production practice, the feedback control system, which is based on the measurement of controlled variables, accounts for an absolutely big proportion. However, there is still a type of situation, process controlled variables, and even process disturbances can not be measured or is difficult to measure. The composition of bottom and top product of a distillation column is the classical example. It is impossible to realize a feed-back control or feed-forward control. The distillation column, as popular separation equipment, is of fundamental importance to the chemical and process industries and its control programs have a very crucial role in the chemical process. To meet the technical requirements and conservation, the composition of top and bottom product stream needs to be controlled at their set values. Theoretically, it is the best to use the composition of the product as the controlled variable, but there are still some problems to limit its usage in practice. Based on the fact that those variables difficult to measure or with a big lag can be inferred from other measurable auxiliary variables, inferential control can compensate for this defect. Taking multi-component distillation process for example, the inferential control scheme is designed and simulated in MATLAB platform. Moreover, its robustness is analyzed. This results reveal that such a scheme achieve an acceptable performance.