Banca de DEFESA: VICTOR SANTOS MATOS

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : VICTOR SANTOS MATOS
DATE: 16/05/2023
TIME: 09:00
LOCAL: https://conferenciaweb.rnp.br/webconf/pse-ufba
TITLE:

Embedded predictive control based on fuzzy model for electrical submersible pump-lifted oil wells

KEY WORDS:

Predictive Control; Fuzzy; Embedded Control; Artificial Method Lift; ESP; Oil and gas.

PAGES: 81
BIG AREA: Engenharias
AREA: Engenharia de Produção
SUMMARY:

Advanced control techniques are not yet a consolidated reality in the oil and natural gas artificial lift sector, mainly in electric submersible pumping (ESP) systems. Model-based predictive controllers (MPC) need accurate and well-representative models, but phenomenological modeling can be expensive and complex due to the dynamics involved. Therefore, data-driven approaches are a robust option to deal with complex phenomena (fluid and thermodynamic) during the production process, in a more economical and effective way. The fuzzy Takagi-Sugeno-Kang (TSK) inference systems is able to approximate the nonlinear dynamics of the elevation systems through a set of linear submodels. Adding this approach to the MPC, it is possible to use a linear control structure and, therefore, numerically-computationally simplified, however, with the appropriate precision and representativeness for the process. Forming a favorable condition for the use in artificial elevation systems, which commonly have isolated installations, dependent on dedicated control systems. The proposal of this work is characterized as the synthesis of a predictive controller with guaranteed feasibility, therefore implementable, with a fuzzy TSK prediction model for ESP-type elevation systems with low computational cost. The focus is the direct application of the technique through the embedding in microcontrollers, simulating the industrial control infrastructure. The case studies implemented on an ESP system through the hardware-in-the-loop approach validated the representativeness of the TSK model and the control capacity and feasibility of the technique, in view of the set of typical restrictions and the reduced number of measured variables.

COMMITTEE MEMBERS:
Presidente - 2042153 - MARCIO ANDRE FERNANDES MARTINS
Externo à Instituição - DANIEL MARTINS LIMA - UFSC
Externo à Instituição - THIAGO PEREIRA DAS CHAGAS - UESC-BA
Externo à Instituição - FLAVIO VASCONCELOS DA SILVA - UNICAMP