Data-reconciliation based fault-tolerant model predictive control for a biomass boiler

Palash Sarkar*, Jukka Kortela, Alexandre Boriouchkine, Elena Zattoni, Sirkka Liisa Jämsä-Jounela

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

3 Citations (Scopus)
148 Downloads (Pure)


This paper presents a novel, effective method to handle critical sensor faults affecting a control system devised to operate a biomass boiler. In particular, the proposed method consists of integrating a data reconciliation algorithm in a model predictive control loop, so as to annihilate the effects of faults occurring in the sensor of the flue gas oxygen concentration, by feeding the controller with the reconciled measurements. Indeed, the oxygen content in flue gas is a key variable in control of biomass boilers due its close connections with both combustion efficiency and polluting emissions. The main benefit of including the data reconciliation algorithm in the loop, as a fault tolerant component, with respect to applying standard fault tolerant methods, is that controller reconfiguration is not required anymore, since the original controller operates on the restored, reliable data. The integrated data reconciliation-model predictive control (MPC) strategy has been validated by running simulations on a specific type of biomass boiler - the KPA Unicon BioGrate boiler.

Original languageEnglish
Article number194
Pages (from-to)1-14
Number of pages14
Issue number2
Publication statusPublished - 2017
MoE publication typeA1 Journal article-refereed


  • BioGrate boiler
  • Data reconciliation
  • Fault-tolerant control
  • Model predictive control

Fingerprint Dive into the research topics of 'Data-reconciliation based fault-tolerant model predictive control for a biomass boiler'. Together they form a unique fingerprint.

  • Equipment

  • Cite this