NOTE: This item is not available outside the Texas A&M University network. Texas A&M affiliated users who are off campus can access the item through NetID and password authentication or by using TAMU VPN. Non-affiliated individuals should request a copy through their local library's interlibrary loan service.
Intelligent fuzzy supervisory control for distillation columns
MetadataShow full item record
Distillation as a separation technique is widely used in the chemical and petroleum industries. With the growth of these industries and the availability of cheap process computing, the interest in more accurate control of the distillation units has increased significantly. The aim of distillation control is to keep the product composition constant in spite of disturbances. Since distillation column is a complex, nonlinear, multivariable interacting system, accurate control is a rather difficult problem. Multivariable design techniques are used to control the column. Some of the different multivariable techniques which have been used in the distillation column control include decoupling control, optimal control and internal model control . Advanced control techniques such as feedforward control are also used for distillation control. The above techniques have been tried in practice and have had varying degrees of success. However, they rely on a valid and accurate model of the process to be controlled. In many difficult process situations such as steelmaking furnaces , cement kilns , presses in the glass industry  and distillation columns , such models do not exist. While there are a variety of adaptive techniques which can partially compensate for this inadequacy , it is clear that if the process cannot be usefully modeled within the framework assumed by the theory, then satisfactory control cannot be achieved. In addition to the model deficiency, a "difficult" process may also be characterized as having a considerable amount of essential a priori information available only in a qualitative form. These features are a form of inexactness or imprecision which prevent the theory from being used [171. Langari and Tomizuka  propose a Fuzzy Linguistic Model(FLM) based feedforward control for improving the disturbance rejection characteristics of a system where an accurate model of the plant is unavailable. For example, in distillation columns, feedforward model and decoupler model are used for tightly controlling the composition in the presence of disturbances and interactions. If these models are not accurate, then feedback controller will be required to take control action. Since feedback controller will not take control action until much damage is done to the controlled variable, tight control of composition is not possible. Because of the changes in the dynamic parameters of the column such as the quality of steam and cooling water etc., a model which is performing well at one operating condition will not perform well at other operating conditions. In practice, the plant is different from the model and there is uncertainty in the plant gain . it is clear that some form of supervisory control is required for dynamically adapting the models to achieve tight composition control. Simple control techniques do not exist for model adaptation in MIMO systems. This thesis will outline a fuzzy supervisory controller based on fuzzy logic and show that control performance can be greatly improved by using such a; fuzzy supervisor. In the proposed fuzzy logic supervisory controller, a set of linguistic control rules will be used to adapt the feedforward model and decoupling model based on the change in the feedflow(disturbance) and the response of the top tray temperature(controlled variable). This thesis will also outline a simulation software to characterize a benzene-toluene binary distillation column and an X-window based Graphical User Interface to run the simulation.
DescriptionDue to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to email@example.com, referencing the URI of the item.
Includes bibliographical references.
Santhanam, Srinivasan (1993). Intelligent fuzzy supervisory control for distillation columns. Master's thesis, Texas A&M University. Available electronically from
Request Open Access
This item and its contents are restricted. If this is your thesis or dissertation, you can make it open-access. This will allow all visitors to view the contents of the thesis.