Abstract
A method is presented for construction of hydraulic characteristics of sections of an oil pipeline by pattern recognition methods. In the theory of pattern recognition, the characteristics of a complex object are studied by means of adaptation algorithms. These algorithms allow the generation of models of processes, establishment of the relationships between their defining parameters and output characteristics on the basis of successive processing of information on the object (for example, dispatchers data on a pipeline sector). The analysis does not require analytic formalization of the processes. This work presents a solution of the problem of determination of the pressure loss to friction over the sections of a main pipeline on the basis of the following initial data: oil flowrate, diameter of pipeline, length of section, and viscosity of oil. The range of change of the pressure drop is divided into intervals (classes), and the task of determination of the continuous value is reduced to recognition of its membership in one of the classes with an accuracy equal to the size of an interval. The potential functions method from pattern recognition theory is used to perform the classification. The algorithm presented allows actual operational characteristics of main oil pipelines to
More>>
Citation Formats
Gubin, V V, Mironenko, N Ya, Titov, N S, and Skovorodnikov, Yu A.
Determination of the actual hydraulic characteristics of a main oil pipeline.
USSR: N. p.,
1976.
Web.
Gubin, V V, Mironenko, N Ya, Titov, N S, & Skovorodnikov, Yu A.
Determination of the actual hydraulic characteristics of a main oil pipeline.
USSR.
Gubin, V V, Mironenko, N Ya, Titov, N S, and Skovorodnikov, Yu A.
1976.
"Determination of the actual hydraulic characteristics of a main oil pipeline."
USSR.
@misc{etde_6215675,
title = {Determination of the actual hydraulic characteristics of a main oil pipeline}
author = {Gubin, V V, Mironenko, N Ya, Titov, N S, and Skovorodnikov, Yu A}
abstractNote = {A method is presented for construction of hydraulic characteristics of sections of an oil pipeline by pattern recognition methods. In the theory of pattern recognition, the characteristics of a complex object are studied by means of adaptation algorithms. These algorithms allow the generation of models of processes, establishment of the relationships between their defining parameters and output characteristics on the basis of successive processing of information on the object (for example, dispatchers data on a pipeline sector). The analysis does not require analytic formalization of the processes. This work presents a solution of the problem of determination of the pressure loss to friction over the sections of a main pipeline on the basis of the following initial data: oil flowrate, diameter of pipeline, length of section, and viscosity of oil. The range of change of the pressure drop is divided into intervals (classes), and the task of determination of the continuous value is reduced to recognition of its membership in one of the classes with an accuracy equal to the size of an interval. The potential functions method from pattern recognition theory is used to perform the classification. The algorithm presented allows actual operational characteristics of main oil pipelines to be defined with an accuracy sufficient for practical application.}
journal = []
volume = {14}
journal type = {AC}
place = {USSR}
year = {1976}
month = {Jan}
}
title = {Determination of the actual hydraulic characteristics of a main oil pipeline}
author = {Gubin, V V, Mironenko, N Ya, Titov, N S, and Skovorodnikov, Yu A}
abstractNote = {A method is presented for construction of hydraulic characteristics of sections of an oil pipeline by pattern recognition methods. In the theory of pattern recognition, the characteristics of a complex object are studied by means of adaptation algorithms. These algorithms allow the generation of models of processes, establishment of the relationships between their defining parameters and output characteristics on the basis of successive processing of information on the object (for example, dispatchers data on a pipeline sector). The analysis does not require analytic formalization of the processes. This work presents a solution of the problem of determination of the pressure loss to friction over the sections of a main pipeline on the basis of the following initial data: oil flowrate, diameter of pipeline, length of section, and viscosity of oil. The range of change of the pressure drop is divided into intervals (classes), and the task of determination of the continuous value is reduced to recognition of its membership in one of the classes with an accuracy equal to the size of an interval. The potential functions method from pattern recognition theory is used to perform the classification. The algorithm presented allows actual operational characteristics of main oil pipelines to be defined with an accuracy sufficient for practical application.}
journal = []
volume = {14}
journal type = {AC}
place = {USSR}
year = {1976}
month = {Jan}
}