Portal de Periódicos da CAPES

Artigo Acesso aberto Produção Nacional

BIBTEX RIS

LPV-H∞ Fault Estimation for Boilers in Sugarcane Processing Plants

2018; Elsevier BV; Volume: 51; Issue: 26 Linguagem: Inglês

10.1016/j.ifacol.2018.11.177

2405-8971

Marcelo Menezes Morato, Paulo Renato da Costa Mendes, Julio E. Normey‐Rico, Carlos Bordons,

Advanced Control Systems Optimization

This short study presents a Linear Parameter Varying (LPV) method for the estimation of faults on Boilers of a Grid-Connected Hybrid Power Plant, that processes sugarcane. These boilers burn residual biomass from the cane and may present a gradual loss of effectiveness, due to internal faults. This renewable energy plant is subject to different processual constraints, having to produce steam, cold water and electric power. The possible faults on the boilers, main source of energy, can lead the system not to comply its operational constraints, which results in direct economic deprivation. These faults are identified through a proposed Fault Detection and Diagnosis approach, based on an LPV extended-state observer, whose gain is derived from Linear Matrix Inequalities (LMI) solutions for the H∞ norm minimization. With the aid of high-fidelity simulation results, the method is discussed.