Intelligent Autonomous System for Electrical Submersible Well Based on Metalearning Techniques

Abstract

Summary This work presents an intelligent autonomous system for Petroleum Processes (SAi2P), whose objective is to monitor and supervise wells for artificial lifting of crude oil specifically by electro-submersible pumping (ESP) integrating operational and financial scenarios. Thus, SAi2P will observe, interpret, make decisions, and execute actions that contribute to maintaining the most efficient operating conditions of the BES, seen from the paradigms of autonomic computing, data analytics (AdD), machine learning, and meta-learning. In general terms, the initial result is a practical Computing System (autonomous sub-cycle of the SAi2P, Analysis of the Operational Behavior of the Oil Well (ACo2P)) and functional that initially allows identifying failure patterns in a concise manner and that makes it possible to carry out simulated tests of the BES operation, which would be unfeasible to carry out in an operational environment.