Project Objectives

The abridged project IdDiaPro searches for ways to diagnose and prognose plant-related or combustion-related, non-nominal operating states  by

(i) model-based analysis methods,

(ii) signal-based analysis methods,

(iii) machine learning methods

and possible combinations of them. The result will be a feasibility study, which includes a review of the methods and their evaluation regarding the implementation in commercial systems during a potential follow-up project.

EIFER’s Contribution

EIFER contributes to the project with its expertise in wood combustion plants for heat production and in signal-based, system-independent diagnosis/prognosis algorithms. The core contributions are:

• Identification of methods for plant-related problems
• Application of signal-based methods for detection of non-nominal states* and diagnosis
• Exploitation of the results

*e.g. more than the triple standard deviation (based on the signal noise) from the system  behavior defined by the system producer