Showing papers by "Azerbaijan State Oil Academy published in 2012"

Decision theory with imprecise probabilities

Abstract: There is an extensive literature on decision making under uncertainty. Unfortunately, up to date there are no valid decision principles. Experimental evidence has repeatedly shown that widely used principle of maximization of expected utility has serious shortcomings. Utility function and nonadditive measures used in nonexpected utility models are mainly considered as real-valued functions whereas in reality decision-relevant information is imprecise and therefore is described in natural language. This applies, in particular, to imprecise probabilities expressed by terms such as likely, unlikely, probable, etc. The principal objective of the paper is the development of computationally effective methods of decision making with imprecise probabilities. We present representation theorems for a nonexpected fuzzy utility function under imprecise probabilities. We develop an effective decision theory when the environment of fuzzy events, fuzzy states, fuzzy relations and fuzzy constraints are characterized by imprecise probabilities. The suggested methodology is applied for a real-life decision-making problem.

Liquid-phase adsorption and isomerization of C5-C8n-paraffins

Abstract: The liquid-phase adsorption of n-pentane, n-hexane, n-heptane, and n-octane from natural gasoline on zeolite CaA and their catalytic isomerization has been investigated experimentally and theoretically with the aim of increasing the octane number of a low-octane gasoline. An integrated process flowsheet combining processes in an adsorber and in an isomerization reactor has been developed. The basic results are as follows: the ultimate activity of CaA with respect to n-pentane, n-hexane, n-heptane, and n-octane in the case of their simultaneous adsorption at 25.0°C is 4.2, 4.7, 5.1, and 6.3 kg/100 kg, respectively. Kinetic and outlet adsorption data are also presented. The maximum yield of C5, C6, C7, and C8iso-paraffins is 62.0, 70.0, 66.0, and 47.0%, respectively. A mathematical model of the processes has been developed, and their parameters have been calculated. Calculated and experimental data are in satisfactory agreement.

Research of a class of smooth membership functions and its application in recognition systems

Abstract: A class of membership functions for fuzzy sets, which generalize known types of membership functions was offered and investigated Was analyzed effectiveness of the hand-printed recognition system using neuro-fuzzy network with membership functions from different classes Analyzed results of numerical experiments of recognition system with membership functions from different classes Revealed comparative advantage of smooth membership functions offered by authors

The analysis of how the choice of membership functions influences the quality of recognition system

Abstract: Was analyzed effectiveness of the hand-printed recognition system using neuro-fuzzy network with membership functions from different classes. For pattern features was constructed smooth membership functions, and also known triangular, trapezoidal, Gaussian and bell membership functions. Analyzed results of numerical experiments of recognition system with membership functions from different classes. Revealed comparative advantage of smooth membership functions offered by authors.

Development of credit risk management system for non-bank credit organizations

Abstract: The management of risk data and information is key to the success of any risk management effort regardless of an organization's size or industry sector Risk management information systems/services (RMIS) are used to support expert advice One of the most important risks that need to manage is credit risk This work is about developing the credit risk management system System should define risks at credit approval level and monitoring interest/principal repayments

Some approaches to solve feedback control problems for nonlinear dynamic systems

Abstract: Optimal feedback control problems for objects described by ordinary differential equations systems on the class of zonal control functions involving uncertain information on the values of these objects' initial conditions and parameters are investigated in the work. Zonal values of the controls are optimized in the problem considered. Formulas for the gradient of the target functional are derived. These formulas allow solving the problem numerically using efficient first order optimization methods. Results of numerical experiments carried out by the example of solution to model problems are given.

Modified Board System for Ecological Monitoring of Ground-Based Objects State

Abstract: In order to improve characteristics and to increase reliability of board systems used at researches of the ground objects, estimation of natural resources and control of ecological state of environment, determination and estimation of accidents and explosions of industrial fields, etc. the structural and methodological solutions are offered. At that increase of system response speed and reliability of received results are provided.

Models and algorithms for a multilevel control system of primary oil refinery installations

Abstract: Algorithms and mathematical models for the technological process of primary oil refinery operating in the uncertain conditions are developed; the solution of the optimal control problem in the form of stochastic programming with probabilistic characteristics is presented. For solving the optimization problem, using the Lagrange method, the problem of development of decomposition algorithms is described and the method based on the transformation of the original problem according to the principle of deterministic analogue is proposed. The construction of the optimal control system created based on the developed models, optimization algorithm, and principles of automatic control of regime parameters of the primary oil refinery installation are considered.

Numerical method of solving the inverse problem of water–oil plug displacement from an oil pool

Abstract: Introduction. It is known that artificially maintaining the pressure in a pool by waterflooding is one of the main technological methods of oil pool development. During flooding, as well as in a natural water-drive regime of development, oil is displaced from a pool by water. It is evident that to predict the performance of an oil deposit after its flooding, it is necessary to use a mathematical model of the process of water–oil displacement from a pool. Experimental and theoretical investigations of various processes proceeding during the development of oil deposits with the use of edge and circle water floodings show that under certain conditions it is sufficient to use the model of plug water–oil displacement to calculate the development. The model of plug displacement presupposes that one liquid is displaced or replaced by another (oil by water) completely and a distinct boundary is formed in the pool between two liquids which in motion obeys the earlier unknown law. It should be noted that efficient development of oil deposits by flooding depends greatly on the regimes of the extraction of oil and water injection into a pool which influence the dynamics of motion of the boundary between the liquids. The rate of motion of the oil–water boundary determines the deadlines of development and flooding, the coefficient of recovery, etc. In this connection, for the practice of the development of oil deposits with the use of flooding, a very important problem is that of regulation of motion of the boundary between two liquids in a pool. This problem is also of current interest for the development of gas fields with active edge or bottom water. The creation and maintenance of underground storages in water-bearing beds and depleted flooded deposits are also associated with the problem of motion of the boundary between two liquids. In the present work, the problem of regulation of motion of the boundary between two liquids is considered as an inverse boundary-value problem for the system of equations of plug displacement of oil from a pool by water. Statement of the Problem. Let there be a homogeneous oil-carrying bed located horizontally of length L and of constant thickness and width. At the boundary x = 0 there is the driving gallery, and at x = L, the operational gallery. At the moment t = 0 water starts being pumped down into the pool through the delivery gallery. In the pool, the oil is plug displaced by water in a rectilinear-parallel manner. It is assumed that the liquids are weakly compressible and that their motion in the pool obeys the Darcy law. Then, the system of equations that describes the process of plug displacement of oil by water in the pool considered can be represented in the following form [1]:

Analysis of the regimes of controlling transient processes in pipelines

Abstract: This paper considers the problem of optimal control of the process of going from one stable regime to another in piping the hydrocarbon raw material We have investigated the influence of the transient period on the dissipation factor of the system, the length of the part of the pipeline, and the difference between the parameters of the initial and final stable regimes with account for the technological limitations set on the control actions and phase variables The dependence of the optimal transient period on the class and range of permissible control functions, the process parameter, and its initial and final regimes has been analyzed

Numerical method of parametrical identification for nonlocal parabolic problems

Abstract: The inverse problem for a parabolic equation with an unknown coefficient on the right-hand side is considered in this work. The dependence identifying coefficient only from variable of time and only from phase coordinates is investigated separately. The numerical method to solution of the problem by using the method of lines is suggested. The results of numerical experiments on test problems are given.

Problem of determination of leakages in oil pipeline networks

Abstract: The problem of determination of breakpoints and of volume of oil leaks under non-stationary regimes of transportation over a linear part of a main pipeline is investigated in the work. The problem of identification is considered in the frame of class of optimal control problems and the formulas for the gradient of the target functional are obtained for the considered problem. These formulas allow using the numerical methods of the first order to solve the problem. The results of numerical experiments are given.

Anomalous diffusion influence on chaotic systems by observation

Abstract: In report's thesis's is interacting systems new structure type Lorenz — Rossler — Anomalous diffusion Obtained the recurrent diagrams, characterizing the features of this structure

Optimal control of a concentrated system on the class of piecewise constant functions under uncertainty in the parameters and initial conditions

Abstract: The authors analyze optimal control problems for objects described by systems of ordinary differential equations on the class of piecewise constant control functions with uncertain initial information about the parameters of the initial conditions and about object parameters. Piecewise constant values of the controls and, what is most important, the boundaries of the intervals of constancy of the controls are optimized in the problem. Given the number of the intervals of constancy, the necessary optimality conditions and formulas for the gradient of the objective functional are obtained. These formulas allow using efficient first-order optimization methods. For the case where the number of intervals of constancy is not specified, an algorithm of its optimization is proposed.

Methods and means for arranging information content searches

Abstract: The rapid growth of the quantity of text documents that are represented in electronic form requires adequate approaches to information content searching to be developed. In order to achieve information content searching, it is proposed, in addition to contextual and paradigmatic means, to use ontology, Semantic Web, and feedback methods with inquiry correction. The comparative analysis of these methods and means is preformed and recommendations on their use are given.

Method of numeral modeling of elastic water drive mode of development of layer

Abstract: We propose a method of numerical modeling elastic water drive mode in the of reservoir development when setting the law of motion of fluids Applying the methods of rectifying the fronts and the difference approximation, the problem reduces to the problem of the difference A computational algorithm for solving the resulting difference problem is given

Indirect indexes of the operational regimes of gaslift wells

Abstract: As author's opinion, there are different important points to get main formulas and thought about output productivity of oil-lift wells. Necessary information such as the instantaneous relative velocity of change of a production rate was given for wells to find suitable way. Moreover, interval of change of a working substance was explained with concrete numbers in the sheet.

Numerical solution to an inverse problem for quasilinear parabolic equation

Abstract: In the work, we consider a process described by a quasilinear parabolic differential equation under some initial and boundary conditions. We set a problem of identifying the heat conductivity coefficient as a function of temperature value. Problems in such a statement arise when studying qualitative properties of materials and media which are not yielding to direct measurements, but studied by means of indirect measurements. To solve the problem formulated, we propose to seek the unknown function (heat conductivity coefficient) on the class of piecewise constant functions. With this end in view, we quantize the set of the phase state's values (temperature) by means of predetermined values. Thus the problem of finding the unknown function is reduced to a problem of determining the finite-dimensional vector. To solve the latter problem, we propose to use first-order numerical minimization methods. With this end in view, we derive formulas for the gradient of the target functional in the space of optimizable parameters.