Sharif Digital Repository

Due to the importance of hepatitis disease, especially hepatitis C, from public health view point, this research aims to investigate and suggest an efficient medical therapy for this disease. First, available models are analyzed and the suitable one is chosen. Three controllers namely, fixed PI, adaptive PI and exact linearization control (GLC) are designed to control the hepatitis C disease using the drug efficacy as the manipulated variabel. Controller parameters are determined by the Genetic Algorithm (GA) and their performances are evaluated using the selected model. Uncertainties have been considerd in both model structure and parameters and controllers performances are investigated in... 

In every year malaria has enormous fatality around the world, so this disease has to be controlled. In this thesis based on a model given in the literature[12], a model predictive controller has been designed and its performance has been compared with that of the optimal control. The simulation results show that in the case of model mismatch, the performance of model predictive control is superior. In model selection, it has been tried that a comprehensive model that has enough inputs for control to be selected. A linear time varying model is used to simulate the nonlinear behavior of the nonlinear system. The Least Squares method is used to update the model parameters  

Consideration of constraints on process variables is an essential part in controller design. In this project, controller design methods for systems with full state constraints were studied and advantages and disadvantages of these methods were described. A controller has been designed for a nonlinear strict-feedback system with unknown dynamics, unavailable states, time-varying asymmetric state constraints, input saturation and unknown control direction. The backsteppping method has been used to design the controller. With the aim of the surface error transformation technique, the problem of full state constraints has been solved. Dynamic surface control has been used to reduce the... 

Immediate control of pandemic influenza to avoid casualties and financial losses is an important issue. In this thesis, two types of controllers including optimal controller and adaptive generalized predictive controller based on linear input-output model, with vaccine, treatment and quarantine being controlling variables, have been used to control influenza. Each controlling methods were investigated for different states including single-input single-output and multi-input single-output. Single-input single-output and multi-input single-output models are identified as online and offline using least squares method. Since it is difficult to exactly identify the model parameters, especially at... 

In this project design of adaptive neural network controller for uncertain nonlinear systems in the presence of different restrictions has been investigated. After providing introduction and motivation, two general forms of nonlinear systems including strict- feedback and pure- feedback have been introduced. In order to solve the problem caused by each restriction, a solution has been proposed. For the strict- feedback nonlinear systems several restrictions are considered such as, unknown dynamic model, unmeasured state, input saturation, unknown control directions, time delays in states and input. Stability analysis based on the Lyapunov theorem has been established. In addition, two kinds... 

This project is aimed to design a controller for nonlinear systems that satisfy a prescribed performance while guaranteeing close-loop stability. To do this, the controller must be designed in such a way that the transient response of the system has had these performance indices and does not exceed them. These performance indices predefine some constraints for systems’ output.A comprehensive form for the model system, the unknown dynamics of the system, consideration of a variety of constraints on the system input, usage of the time-finite method for the residence time and consideration of just one tuning parameter to reduce the computation load, are among the other benefits for the proposed... 

The Smith Predictor does not work for unstable plants; therefore a lot of modifications have been proposed in order to solve this problem. In this project, advantages and disadvantages of most of these methods are studied in a general framework and based on Diophantine equation and using pole-placement technique. a discrete control scheme for unstable processes has been proposed. This structure can be used for non-minimum phase and stochastic processes. Robust stability of the suggested control structure has been shown. Performance of the proposed method is compared with previous works through computer simulations and its superiority for robust stability, load rejection; set-point tracking... 

Control of fermentation bioreactors with nonlinear behavior is a very important issue, especially when optimizing the biomass production is desired. The system in this situation works in two domains: minimum phase and non-minimum phase; and in transition from one domain into the other one the sign of process gain changes. For control of such a process in this work have been used two non-adaptive strategies and one adaptive strategy. The non-adaptive schemes are pole-placement and model predictive. The pole-placement technique as a linear control algorithm is used and its parameters are fixed by minimizing the objective function which has been used in predictive control technique. Another... 

Sudden pressure drop and then the gas shortage in city gas stations is one of the most important problems in high-pressure natural gas transmission networks. On the other hand, power consumption of compressors imposes a lot of cost on compressor stations. The conventional schemes to navigate the compressor stations are traditional methods based on experiment and trial and error which cause energy waste and violation of the optimal operating conditions. In this research, two manners of centralized- and decentralized- control are employed in order to navigate compressor stations in specified operating conditions regarding the operational and practical constraints. In this purpose, a sample gas... 

Most of Final properties of polymers such as morphological characteristics are related to molecular weight distribution (MWD) and particle size distribution (PSD). It is possible to obtain the desired properties by controlling of these indices. In this work, modeling, simulation, and control of final properties of styrene have been considered. First, kinetic of reaction has been investigated which is a free radical type. Molecular weight distribution has been calculated by using moment method. Next, population balance equation (PBE) has been used to model PSD. For this system, mass, energy, and moment conservation equations are a set of stiff ordinary differential equations (ODEs). PBE is a... 

Obtaining polymers with desired properties is very important from industrial point of view. Weight average molecular weight and polydispersity are two important polymer properties. It is also important to obtain a high monomer conversion. In this thesis production of polyvinyl acetate is considered. Under special conditions, the vinyl acetate reactor dynamics become oscillatory which makes the reactor control difficult. For obtaining polymer with desired properties, an objective function which includes weight average molecular weight, polydispersity and monomer conversion has been defined. Then by minimizing this objective function the optimal temperature trajectory is obtained. In this... 

In this research, it was tried to present a suitable environment for simulation of crystallization systems. Chapter one presented a survey on crystallization systems, its application, operating methods, and its phenomena. Chapter two presented mechanism and kinetic of phenomena that occur in crystallization systems, in detail, and mathematical models to modeling this systems are implied. In chapter three, that is the simulation result section, firstly, it focused on solution methods to solve population balance equation, that explains the particle size distribution, and its validation, secondly, the simulation results in a case study was compared with those obtained from finite volume method.... 

In this work, a comprehensive dynamic model has been used for vinyl acetate emulsion polymerization in a continuous reactor to predict the evolution of product particle size distribution under isothermal condition. Method of finite-volume has been applied for solving the population balance equations and the results are compared with the results obtained from moment method. Finite-volume method has been selected as a precise technique to predict sustained oscillations, which occurs in continuous emulsion polymerization. After performing sensitivity analysis and verification of system’s controllability, feed rate of surfactant and initiator have been selected as proper manipulated variables to... 

In this work, a comprehensive dynamic model has been used for styrene emulsion polymerization in a semi-batch reactor to predict the evolution of product particle size and molecular weight distributions under non-isothermal condition. Method of moments has been applied for solving the population balance equations and the results are compared with the results obtained from finite-volume method as a precise technique. A sensitivity analysis respect to perturbation in initial values of various reagents, species and the reactor temperature has been performed to identify suitable manipulated variables for closed-loop control of the reactor. The detailed model is used to generate optimal feed... 

This thesis concerns the modeling of precipitation processes aimed at predicting product particle characteristics especially particle size distribution. A mixing-precipitation model (based on SFM) linked to the population balance is used for a single-feed semi-batch precipitation process to describe the mixing effects on the final particle-size distribution (PSD) and mean size, considering the agglomeration mechanism. The model describes mixing on two zones: the feed and bulk zone with different levels of super saturation. To solve a full population balance equation containing nucleation, growth and agglomeration, discretization method was performed witch is more accurate and low... 

In control systems, actuators are subjected to faults during operation. In the present work, a controller for nonlinear systems under several restrictions and in the presence of actuator failure has been proposed. The controller is designed for large-scale multi-input multi-output (MIMO) uncertain nonlinear systems subject to unmodeled dynamics, unknown state dependent control gain functions, unknown state time varying delay, unknown subsystems interactions, unknown external disturbances and input nonlinearity. The backlash has been assumed for input nonlinearity. No restriction has been considered regarding the number of actuator failure. The backstepping technique has been applied for... 

This thesis focuses on modeling, simulation and control of bulk polymerization of styrene in an autorefrigerated reactor where the generated heat is removed by heat of vaporization of the monomer. By means of two mathematical models for the kinetics of reaction, static and dynamic simulations were performed. For this purpose the equilibrium points were calculated using the steady state analysis, and then the reactor temperature was controlled at the unstable point using PI and adaptive generalized predictive controller (AGPC). In order to implement the GPC scheme a process model is required to provide the process output estimate over the prediction horizon. For this purpose an auto... 

Suspension polymerization is an important industrial process. Product of this process is particulate polymers that have a widely use in human lives. An important property of suspension polymerization process is the particle size distribution (PSD), which affects the end-use properties of the products. Therefore, the evolution of PSD is a major concern in the design of a suspension polymerization process. In a suspension polymerization reactor the evaluation of particle size distribution is the result of the interaction of the mixing conditions and the kinetics of the polymerization reactions. Mathematical modeling of particle size distribution in suspension polymerization processes should... 

The diesel fuel produced from vegetable oils or animal fats is called biodoesel. Biodiesel has desired fuel property as an alternative for diesel engines. Moreover, It is biodegerable, non-toxic, environmental friendly and has lesser undesirable exhaust. Due to biodiesel advantages, production of biodiesel and control of the biodiesel production reactor is noteworthy. The objective of this work is to control the mass fraction of the produced ester and the reactor temperature by manipulating methanol feed flow rate and the heat duty of the reactor. For this purpose, an adaptive multivariable generalized model predictive control (GPC) has been applied to the reactor. The results indicate that... 

This thesis addresses design of an adaptive tracking control of input-quantized strict-feedback fractional-order nonlinear systems with unknown control directions. The control design is achieved by using a hysteretic quantizer to avoid chattering. The main advantages of the proposed controller are: the fractional order is available for both commensurate and non-commensurate cases, the¬¬ controller design does not depend on the quantization parameters and some restrictive assumptions are relaxed. Additionally only one adaptive law has been used for the unknown control directions which leads to reduction of computational load. By utilizing the backstepping approach and based on the frequency...