Sharif Digital Repository

Data clustering, is one of the most important problems in computer science. It has countless applica-tions in many different topics like machine learning, image processing, computer vision and bioinfor-matics. However, current algorithms for this problem do not have high accuracy and speed. In this thesis, we have proposed a parallel algorithm for data clustering based on the solution of Isoperimetric problem on trees. In addition, we have implemented the algorithm on GPU as a powerful, cheap and mostly available hardware. Finally, by doing different practical experiments, we have shown that the algorithm has more than 20x speed-up over the sequential algorithm without changing in the... 

Many spacecrafts and satellites have long flexible components, the most common ones are the satellites with flexible solar panels, which are used to provide the satellite energy. The governing equations of the satellite rigid motion along with the solar panels vibrations are the combination of the coupled partial and ordinary differential equations. Various reasons, such as the satellite maneuvers and the external disturbances, cause the vibration of panels. Due to the dependence between the governing equations of the satellite rigid motion and the flexible components vibrations, these vibrations affect the satellite rigid dynamics and the control aims; therefore, in addition to the... 

Nowadays, micro systems in many area of science and technology have special importance and status. The main function of these systems is based on the deformation of a beam at micro scale. Therefore study of behavior and control of micro-beams will have a great importance. Among micro-beams, clamped-free and clamped-clamped micro beams that actuated by piezoelectric and electrostatic actuators have wider applications. For example, clamped-free micro-beams are used in atomic force microscopy (AFM), micro switches, mass sensors and micro-accelerometers and clamped-clamped micro-beams are used in micro mirrors and Grating Light Valves (GLV). In this thesis, the studied system is clamped-free... 

Gas transmission networks play an important role in multi-vector energy networks. Therefore, reliable and efficient control of gas transmission networks with the minimum energy consumption is utmost task and is still an active research area. This work, introduces a distributed, model-free, and game-theoretic approach comprising a simultaneous game followed by a sequential game, to achieve the control objectives in gas transmission networks. The cooperative game approach is adopted for this purpose. In the proposed method two fuzzy inference systems have been used to obtain the payoff of each player in the game of gas transmission network control, the first one has been used to determine the... 

Natural Gas is one of the most important energy resources in the world. The U.S energy information administration(EIA) has reported that natural gas consumption will increase until 2030 in all participating sectors including industrial, domestic and electricity production.Although the use of natural gas as a source of energy has many advantages, natural gas physical condition made it’s transfer so challenging. Among the various types of transmission methods, using pipline is so conventional. Natural gas can be transmited through pipelines in a safe and reliable manner upto thousands of kilometers. For this reason, operation monitoring and control of gas transmission networks are so... 

Identification and control of nonlinear systems have been always such important subjects based on nonlinearities of many processes. For instance, identification and control of gas transmission networks can be mentioned as one of the challenging problems related to their complexity and existence of different variables. The purpose of this project is identification of a nonlinear model for defining a gas transmission network and implementing a model predictive control algorithm based on identified model. Hence, first, a volterra model is implemented for identification of a pH neutralization process and then, that model is used for identification of a gas transmission network. Finally, by... 

Due to continuous improvements in manufacturing and the fact that modern technologies such as microprocessors, are becoming more efficient along the rapid growth of nonlinear control methods usage, the application of these two tools together has entered various fields of biomedical engineering, particularly neural prostheses. In this study, the methods of analysis and control of nonlinear systems were used to investigate and improve the behavior of the neural breathing system on the basis of Hodkgin-Huxley model as the original model of the neuron. Using the principles of differential geometry and robust control methods – based on combination of exact input-output linearization ideas,... 

In this thesis, using superstructure concept, a method is presented to synthesize of new structures for thermal desalination processes along with their structural and operational optimization. The proposed superstructure is capable of producing various common (feed forward, backward, parallel cross feed) and non-common structures in evaporators, as well as in combination with flash stages and thermal vapor compressor. By using the GA (Genetic algorithm), the superstructure has been optimized to identify the optimal structures in both Iranian and international economic scales. Also comparisons were made between conventional structures at optimum state, to suggest superior structures at... 

The ratio of the two first non-trivial Dirichlet eigenvalues is the subject of the Payne-Polya-Weinburger Conjecture and Inequality. The PPW Conjecture was proven by M. Ashbaugh and R. Benguria. Ashbaugh and Benguria later proved a similar result for domains on the n-sphere. On the other hand, discrete counterparts of universal inequalities from (continuous) have been studied in the graph theory setting. This thesis surveys results related to the PPW inequality on graphs  

Diabetes is one of the most epidemic metabolic disease which needs to be controlled through medications. The main objective of this study is to model and control diabetes type1 with exercise. It is important to choose an appropriate yet simple model based on which the design of the controller is accomplished (Bergman model is the one which is appropriate for this purpose). On the other hand, a comprehensive model which is used as the virtual patient has been chosen to assess the performance of the controller designed based on simple mode. The chosen comprehensive model is Cinar’s model. A good control of diabetes was achieved when the glucose blood (GB) of the two models almost had the same... 

Natural tissues of body consist of Extra Cellular Matrix (ECM) and cells. Scaffold produces a temporary matrix for cells and facilitates oxygen and nutrients transport which enhances tissue regeneration. Therefore, attempts in regeneration of tissues and organs by scaffolds attracted great interest in today’s researchs. Choosing scaffold fabrication method and its polymer type, play a key role in scaffold specifications as they determine scaffold properties. In this project electrospinning was chosen as the scaffold fabrication method which is because of the unique properties of scaffolds fabricated by this method and also its ability in producing micro and nano fibers. At first set of... 

Steam-Assisted Gravity Drainage (SAGD) is one of the best and the most effective methods of enhanced oil recovery (EOR) which is used in the production from Heavy Oil Reservoirs which constitute large amount of oil throughout the globe. In a typical process of SAGD, steam is injected into a horizontal well and the oil is produced several meters deeper in the ground. In order to increase the Recovery of this process, it is better to review on the Pre-heating methods around the injecting and producing wells. Pre-heating is important when the viscosity of the oil is large enough to prevent oil from any motion at the beginning of the SADG operation, and makes it necessary to to use a Pre-heating... 

Management decisions, enhanced oil recovery, and reservoir development plans in petroleum industries are based on predictions by reservoir simulation. Due to uncertainties in model parameters or engineering assumptions, the simulation results are not accurate, while they are correct. For more accurate estimation of unknown production quantities, it is required to characterize the unknown parameters and its uncertainty. By using static data alone the result of characterization is unreliable and unsure, therefore dynamic data use practically. In reservoir engineering literature, this is called “History Matching”.The ensemble Kalman filter is an optimal recursive data processing algorithm based... 

Hydrocarbon reservoirs development and management is based on their dynamic models. To encounter various types of error during model building, model parameters are adjusted to produce reservoir historical data by assimilation (history matching) of reservoir production or 4D seismic data. Among the existing sequential methods for automatic history matching, ensemble Kalman filter and its variants have displayed promising results. The innovations of this thesis for ensemble Kalman filter (EnKF) are presented into three major orients; these includes adaptive localization/regularization, characterization of original PUNQ test model and characterization of channelized reservoir.
To mitigate... 

In this research, the aim is to investigate the use of data assimilation method for better reservoir model updating in the reservoir management. In addition, multi-objective optimization concept is studied for the production optimization. The application of these methods is applied for reservoir management in the naturally fractured reservoirs. To update the reservoir, ensemble based methods, especially ensemble Kalman filter and ensemble smoother, are used. To tackle the challenges encountered with these methods, modifications are proposed to obtain a better history matching and more accurate reservoir characterization. The proposed framework for history matching is implemented for the... 

A modeling framework is proposed for complex chemical process networks. Concept such as passivity systems, thermodynamic function and graph theory are some important and applicable concept which seems to be useful for process network. The state is represented by energy, volume and mass inventories. The dynamic behavior of the process system is constrained so that all trajectories satisfy the first and second laws of thermodynamics. The concavity of the entropy function is used to define a storage function for passivity design. The proposed storage function is closely related to the Gibbs tangent plane condition. Stability analysis, pairing and controller design could be done by this new... 

Fuel cell-based CHP systems for distributed residential power generation represent an interesting alternative to traditional thermoelectric plants. This is mainly due to the high efficiency obtainable in the production of electricity and heat in a decentralized, quiet and environmental friendly way. The current thesis focuses on the development, in Matlab_Simulink environment, of a complete dynamic model of a residential cogenerative (CHP) energy system consisting of the Proton Exchange Membrane fuel cell (PEMFC), in two scenarios. The first scenario emphasizes on use of fan and usual Heat Exchanger for cooling of Proton membrane fuel cell and conveying generated heat in PEM fuel cell to... 

Whenever continuous and discrete dynamics interact, hybrid systems arise. Basically, models of such systems consisted of differential or difference equations to represent continuous dynamics and automata or other discrete variables to describe behavioral mode of the systems. Our main concern is focused on two types of hybrid systems: Discrete controllers which are combined with continuous physical process like biological systems can often be well described by hybrid systems. Our approach in this thesis deals with optimal multiple daily insulin injections which are introduced as an offline control approach on a simulated patient with type I diabetes. Simulated patient is provided via GlucoSim... 

A mathematical model helps to understand any system with any level of complication, such as the human body. The heart, vascular system, and lungs are three important parts of each body that life is impossible without them. The performance of the cardiovascular and respiratory systems is tuned by a group of complex mechanisms which are called regulatory mechanisms. Therefore, a model which describes the behavior of these mechanisms has great importance. The goal of this research is to provide and simulate the newest and most complete model in this field. The presented model should consist of the heart, vascular system, lungs, respiratory system, gas exchange and transport, and finally... 

In this study, single-phase and two-phase, incompressible flow in a non-deformable porous media was modeled and also simulated. The Reynolds numbers have been considered high enough to justify the use of the model including inertial terms. Afterwards, the mass conservative equations were discretized. Then, the numerical model has been developed for heterogeneous porous media. The heterogeneous porous media was comprised by several specified homogeneous regions, each having isotropic transport properties or permeability distribution with applying stochastic methods. The developed numerical model was used for predicting of saturation profile in water flooding process in one, two and...