-
Rodrigo L. Meira
Programa de Pós–Graduação em Engenharia Industrial (PEI), Escola Politécnica, Universidade Federal da Bahia, Rua Prof. Aristides Novis, 2, Federação, 40210630, Salvador (BA)
-
Márcio A.F. Martins
Programa de Pós–Graduação em Engenharia Industrial (PEI), Escola Politécnica, Universidade Federal da Bahia, Rua Prof. Aristides Novis, 2, Federação, 40210630, Salvador (BA)
-
Ricardo A. Kalid
Programa de Pós–Graduação em Engenharia Industrial (PEI), Escola Politécnica, Universidade Federal da Bahia, Rua Prof. Aristides Novis, 2, Federação, 40210630, Salvador (BA)
-
Gloria M. N. Costa
Programa de Pós–Graduação em Engenharia Industrial (PEI), Escola Politécnica, Universidade Federal da Bahia, Rua Prof. Aristides Novis, 2, Federação, 40210630, Salvador (BA)
Keywords:
Infinite-horizon control, surge avoidance, natural gas network, MPC, adaptive control
Abstract
This work proposes a practical NMPC-based model predictive control approach with surge avoidance to natural gas networks. A minimal order state space-type model formulation based on an analytical expression of the step response has been proposed by linearizing the partial differential equations (PDE) for pipe flow. The simulation results showed that the proposed controller succeed in tracking the reference points meanwhile avoiding to remain on risk conditions, violating the surge line constraints. The closed-loop computational cost is also compatible with the pipeline network dynamic with an average time of around 2.5 s. Such results present the potential of the proposed approach compared to literature techniques such as simplifications on PDE or data-driven models.
