ОЦЕНКА ТЕХНИЧЕСКОГО СОСТОЯНИЯ УСТАНОВОКЭЛЕКТРОЦЕНТРОБЕЖНЫХ НАСОСОВ В ПРОЦЕССЕ ЭКСПЛУАТАЦИИ МЕТОДОМ НЕЙРОСЕТЕВОЙ КЛАССИФИКАЦИИ

Abstract

Nowadays electric submersible pumps (ESPs) are widely distributed and that’s why attention is paid to work out solutions on effective control of the active well stock via technical condition evaluation of down hole equipment while operation. There is a possibility of equipment’s failure because of its complex design. But there are ways to reduce it. Quality and efficiency of decision on technical condition of equipment is largely depends on employees background who works in the operation segment. Fluctuation of electrical parameters of ESP during its operation is in the stochastic nature. And there is huge amount of technological data indicating the operating mode of ESP and what affects on error probability of incorrect technical condition evaluation of the equipment and a wrong decision on the operating mode. This paper proposes a method of diagnosing the condition of ESP, based on the using of a device with software for analysis of stochastic technological parameters in timescale form, which approximated with statistical evaluation with the neural network algorithm classification. That ways are implemented in a device which allows evaluation technical condition of pump during its operation with high veracity. In this way the problem of high accuracy of technical condition evaluation of ESP solves.