Research Interests:
New stability constraints are developed for sequential stabilization of decentralized control systems. Contrary to previous results, the current methods are derived directly from the stability condition of the whole decentralized control... more
New stability constraints are developed for sequential stabilization of decentralized control systems. Contrary to previous results, the current methods are derived directly from the stability condition of the whole decentralized control system. These results are based on the Structured Singular Value and lend themselves naturally to a sequential stabilization procedure as well as the criteria to select a priori the order of the controller design sequence.
Research Interests:
This paper presents a MIMO nonlinear predictive control technique which can handle processes with input and output constraints. The method is a model based control strategy which uses a nonlinear dynamic model in the control loop. Control... more
This paper presents a MIMO nonlinear predictive control technique which can handle processes with input and output constraints. The method is a model based control strategy which uses a nonlinear dynamic model in the control loop. Control actions are generated from a special iterative QDMC (Quadratic Dynamic Matrix Control) algorithm and the basic features of the classical QDMC are retained. Update of the linear QDMC model by a disturbance vector which accounts for the nonlinearities in the prediction horizon is the novel feature of the method. It is this disturbance vector that drives the proposed nonlinear control algorithm.
Research Interests:
Research Interests:
... The polynomial ARMA models have been used by several authors for modelling and control of nonlinear systems (eg Korenberg et ale 1988,Hernandez 1992,Hernandez and Arkun 1993).Theoretical justifications for this structure are also... more
... The polynomial ARMA models have been used by several authors for modelling and control of nonlinear systems (eg Korenberg et ale 1988,Hernandez 1992,Hernandez and Arkun 1993).Theoretical justifications for this structure are also available (eg see Diaz and ...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests: Engineering, Chemical Engineering, Mathematics, Computational Biology, Nonlinear dynamics, and 15 moreNetworks, Medicine, Multidisciplinary, Biological Sciences, Humans, Hypertension, Insulin, Diabetes mellitus, Interactions, Renin Angiotensin Aldosterone System, Ii, PLoS one, Implications, Angiotensin II, and Angiotensin
Abstract A conceptual framework to design robust process control systems is develope d and its realization through an interactive computer-aided design software is presented. The overall design methodology is based on a unified treatment... more
Abstract A conceptual framework to design robust process control systems is develope d and its realization through an interactive computer-aided design software is presented. The overall design methodology is based on a unified treatment of recent theoretical results in modern control and new computational techniques in symbolic logic manipulation, singular value decomposition and optimization. Several physical examples are given to demonstrate the application of the design approach and the utility of its computer software.
Research Interests:
Research Interests:
Abstract This paper presents for the first time a theoretical nonisothermal dynamic model for spouted bed catalytic reactors. A pseudo-homogeneous on-dimensional model is developed for an adiabatic spouted-bed with a first order reaction.... more
Abstract This paper presents for the first time a theoretical nonisothermal dynamic model for spouted bed catalytic reactors. A pseudo-homogeneous on-dimensional model is developed for an adiabatic spouted-bed with a first order reaction. The effects of various design parameters on the steady-state reactor performance are established.
Research Interests:
Research Interests:
<p>(<b>A</b>) Steady-state response curves of pAKT vs insulin for different feedback strengths . . The curves are calculated by setting <a... more
<p>(<b>A</b>) Steady-state response curves of pAKT vs insulin for different feedback strengths . . The curves are calculated by setting <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0083640#pone.0083640.e030" target="_blank">equations (10</a>) and (11) equal to zero. (<b>B</b>) Dynamic response curve in red superimposed on the steady-state curve shows the normal insulin cycle between states A and B. At State A, the cell has low nutrient level and requires glucose uptake. By stimulating insulin, the system switches to State B, where pAKT is activated and glucose is taken into the cell. Withdrawing insulin enables the switch back to low pAKT levels. (<b>C</b>) Bistability is lost under excess negative feedback Steady-state curves in (A) are identical to those in <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0083640#pone.0083640-Wang1" target="_blank">[45]</a>. Dynamic responses in (B) and (C) are new and they are generated form the dynamic model <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0083640#pone.0083640.e030" target="_blank">equations (10</a>) and (11).</p
Research Interests:
Research Interests:
Research Interests:
Research Interests:
In this paper a new nonlinear model predictive control law is applied to a semi-batch polymerization reactor. The particular model used in this study is the free radical polymerization of polymethylmethacrylate. The principal... more
In this paper a new nonlinear model predictive control law is applied to a semi-batch polymerization reactor. The particular model used in this study is the free radical polymerization of polymethylmethacrylate. The principal nonlinearities are due to the reaction rates and the gel effect. Two cases are considered: jacket temperature controlling the reaction temperature, and a MIMO structure where both
Research Interests:
... Mn-Sen Chiu and Yaxan Arkun School of Chemical Engineering Georgia Institute of Technology, Atlanta, Georgia 30332-0100 ... if (a) Dr is square, (b) det(Dw) E V. and (c) P = NrDg, 3. The pa (Np,,, Dp,r) E M(S) is caled a rikt-prime... more
... Mn-Sen Chiu and Yaxan Arkun School of Chemical Engineering Georgia Institute of Technology, Atlanta, Georgia 30332-0100 ... if (a) Dr is square, (b) det(Dw) E V. and (c) P = NrDg, 3. The pa (Np,,, Dp,r) E M(S) is caled a rikt-prime fratio representtion( rcfr) of PeM(R) ifandoalyif ...
Research Interests:
... mapping operator called the control law. There exist almost infinitely many -alternative con-trollers since measurements can be mapped to mani-pulated variables via different control laws. These range from the simplest single ...
Research Interests:
Research Interests:
This paper discusses the use of the structured singular value (SSV) for robustness analysis of process control systems and presents some new results for distillation columns.
Research Interests:
Research Interests:
Research Interests:
We (1999, 2000) have introduced a model predictive control (MPC) algorithm for polytopic linear parameter varying systems. The algorithm, called quasi-min-max MPC, separates the first stage cost from the infinite horizon quadratic... more
We (1999, 2000) have introduced a model predictive control (MPC) algorithm for polytopic linear parameter varying systems. The algorithm, called quasi-min-max MPC, separates the first stage cost from the infinite horizon quadratic objective function and bounds the rest of the cost by a quadratic term. For predictions the future state matrices are assumed to belong to a fixed polytope. In
Research Interests: Chemical Engineering, Mathematics, Model Predictive Control, Mathematical Programming, Stability, and 12 moreSampling methods, Semidefinite Programming, Predictive models, Linear Parameter Varying Systems, Linear Systems, Upper Bound, Cost Function, Predictive Control, Polytope, Objective function, Infinite Horizon, and Rate of Change
Modeling and Control of an Industrial Hydrocracker Using the Dicrete and Continuous Lumping Methods and Model Predictive Control Hasan Sildir, Ummuhan Canan,Yaman Arkun. Department of Chemical and Biological Engineering, Koc University,... more
Modeling and Control of an Industrial Hydrocracker Using the Dicrete and Continuous Lumping Methods and Model Predictive Control Hasan Sildir, Ummuhan Canan,Yaman Arkun. Department of Chemical and Biological Engineering, Koc University, Istanbul Turkey. Berna Cakal, Dila Gokce, Emre Kuzu,TUPRAS Refineries, Korfez-Kocaeli,Turkey Hydrocracking is a catalytic chemical process which converts high-boiling heavy petroleum fractions such as vacuum gas oil into lighter and more valuable products like naphta, diesel, kerosene, gasoline, and LPG. Hydrocracking takes place in the presence of rich hydrogen at elevated temperatures and pressures. The products are free of sulphur and nitrogen compunds which are hydrogenated into hydrogen sulfide and ammonia and which are subsequently removed. The aim of this work is to develop a model for an industrial hydrocracking reactor for optimization and control purposes. Our research is centered around the hydrocraking unit (HCU) of TUPRAS refineries whic...
Hydrocracking is a catalytic chemical process which converts highboiling hydrocarbons to more valuable lighter products like diesel, kerosene, naphtha and LPG. Environmental concerns and increased demand for low-sulphur diesel and high... more
Hydrocracking is a catalytic chemical process which converts highboiling hydrocarbons to more valuable lighter products like diesel, kerosene, naphtha and LPG. Environmental concerns and increased demand for low-sulphur diesel and high smoke point jet fuel have played a major role in the increased utilization of the hydrocracking technology. Hydrocracking is carried out in the presence of hydrogen at high temperatures (260 495C) and pressures (35 200 bar). The main reactions are cracking and hydrogenation which take place in multiple catalytic beds. The overall system is highly exothermic and hydrogen quench is used for interstage cooling. Due to hydrogenation, sulfur and nitrogen compounds are removed in the process to yield lowimpurity reactor product. In the fractionation part of the plant, the reactor effluent is separated into the final valuable products. Our research covers the real time plant-wide optimization of the hydrocracking reactors and the fractionation columns of TUP...
Research Interests:
This paper presents a self-scheduling MPC framework for plants described by LPV (Linear Parameter Varying) models. Such a controller adjusts to variations in plant dynamics by using measured values of the parameters in the control law. We... more
This paper presents a self-scheduling MPC framework for plants described by LPV (Linear Parameter Varying) models. Such a controller adjusts to variations in plant dynamics by using measured values of the parameters in the control law. We apply the method to control of nonlinear plants approximated by LPV models constructed from multiple local linear models. In this context the parameters constitute model validity functions which are estimated on-line and used for scheduling MPC. Both quadratic programming based finite horizon MPC and min-max type LMI based MPC algorithms are discussed and applied to continuous and batch systems.
Research Interests:
Research Interests:
Research Interests:
Interaction forces among residue pairs are determined from optimum folding pathways along which a protein represented as a coarse-grained chain of alpha-carbons goes from different initial configurations to a known native state. A dynamic... more
Interaction forces among residue pairs are determined from optimum folding pathways along which a protein represented as a coarse-grained chain of alpha-carbons goes from different initial configurations to a known native state. A dynamic optimization approach is employed that uses the coarse-grained model to compute the optimal folding pathways. The pair-wise interaction forces obtained in this manner are incorporated into the coarse-grained model which is then simulated to fold the protein from a new set of initial configurations in a predictive way. We show that the folding pathways predicted in this manner are near-optimal. We applied the technique to the secondary structures: helix and beta-sheet.
Research Interests:
Research Interests: Engineering, Algorithms, Biophysics, Chemistry, Protein Folding, and 15 moreComputational Biology, Dynamic Optimization, Medicine, Biological Sciences, Physical Biology, Animals, Physical sciences, Peptides, Proteins, Chickens, Equation of Motion, Protein Conformation, PLANT PROTEINS, Protein Biosynthesis, and Energy minimization
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
When a plant-wide control structure is being designed, there are often more measurements and manipulations than are required. Therefore the designer must judiciously select a subset of them that will be used for feedback control. If the... more
When a plant-wide control structure is being designed, there are often more measurements and manipulations than are required. Therefore the designer must judiciously select a subset of them that will be used for feedback control. If the control structure is to be ...
Research Interests:
New stability constraints are developed for sequential stabilization of decentralized control systems. Contrary to previous results, the current methods are derived directly from the stability condition of the whole decentralized control... more
New stability constraints are developed for sequential stabilization of decentralized control systems. Contrary to previous results, the current methods are derived directly from the stability condition of the whole decentralized control system. These results are based on the Structured Singular Value and lend themselves naturally to a sequential stabilization procedure as well as the criteria to select a priori the order of the controller design sequence.
Research Interests:
The discussion includes the theoretical foundations necessary in the synthesis of optimizing control structures for a given plant; the Kuhn-Tucker theory of nonlinear programming and search strategies for the optimizing controller;... more
The discussion includes the theoretical foundations necessary in the synthesis of optimizing control structures for a given plant; the Kuhn-Tucker theory of nonlinear programming and search strategies for the optimizing controller; analysis of large-scale systems composed of various interconnected subsystems; alternative practical operation strategies for the on-line implementation of the optimizing controllers for a chemical plant; guidelines for screening alternative strategies to arrive at an acceptable sequence of optimization steps; the use of steady-state optimizing control strategies on a continuous stirred tank reactor and a fluid catalytic cracking process; current studies on the application of the strategies to distillation columns and a large-scale gasoline polymerization plant; and theoretical studies underway to develop adaptive control strategies to cope with model inaccuracies and their effects on the optimizing control strategies.
Research Interests:
Research Interests:
The major criticism related to the use of the Block Relative Gain (BRG) (and for that matter The Relative Gain) as a closed-loop interaction measure is the fact that it is developed under the assumption of perfect control in the... more
The major criticism related to the use of the Block Relative Gain (BRG) (and for that matter The Relative Gain) as a closed-loop interaction measure is the fact that it is developed under the assumption of perfect control in the complementary subsystem(s).
Research Interests:
In this paper, we consider the problem of designing a stabilizing linear controller for a feedback linearized nonlinear system under the influence of bounded perturbations. The feedback linearized (and perturbed) nonlinear system is... more
In this paper, we consider the problem of designing a stabilizing linear controller for a feedback linearized nonlinear system under the influence of bounded perturbations. The feedback linearized (and perturbed) nonlinear system is equivalent to a Quasi Linear System. A sufficient condition using Lyapunov stability theory is presented to obtain a linear state feedback controller stabilizing the Quasi Linear System and the nonlinear system as well. A method is given to obtain th least conservative (smallest) stabilizing gains. The results are applied on a chemical reactor.
Research Interests:
In this paper new dynamic interaction measures are presented to study the stability and performance characteristics of decentralized control structures. The measures provide different valuable insights into the general control structure... more
In this paper new dynamic interaction measures are presented to study the stability and performance characteristics of decentralized control structures. The measures provide different valuable insights into the general control structure synthesis problem.
Research Interests:
In this paper a new approach to control full CD profiles in sheet forming processes is presented. The method consists of a sequence of dimension reductions of the input, output and external disturbance spaces, while alleviating... more
In this paper a new approach to control full CD profiles in sheet forming processes is presented. The method consists of a sequence of dimension reductions of the input, output and external disturbance spaces, while alleviating ill-conditioning problems. SVD is used to initially condition the input-output space, and then the order of the subspace of the dominant disturbances is identified on-line through the use of the Karhunen-Loeve expansion. Subsequently the dimension of the input space is further reduced by projecting it onto the disturbance subspace. Finally, model predictive control able of handling input constraints is implemented on this space.
Research Interests:
For large sle industrial decentraized cmtrol is preferred because multivariable centralized control reuir where too many control loops with incaeld cost and ty of desg, and difficult plem tat t and maintenance problems. Two dase of desip... more
For large sle industrial decentraized cmtrol is preferred because multivariable centralized control reuir where too many control loops with incaeld cost and ty of desg, and difficult plem tat t and maintenance problems. Two dase of desip procedures for decen The ti tralized controllers have been reported in the literature. The first clas is independent desig whee eac contrller element is desiged independently of each other [12]. The nice feature of such an approach is that each controler is desdiged S diecty. However, it is potentially conservative as pointed st)cand out in [2]. The second clas is sequential desin in which controller desip is conducted sequtiaIly [3,4,5]. At each design t one utilizes the information about the contes M n Re specified in the previous step; therefore, the method ca be T lss conservative than the independent design. nyquist sbifity In this work a new stability constraint for each conterolere a in sequential desip procedure is proposed. The result is ba...