Designing automated fault detection and diagnosis (FDD) mechanism is key to the widespread development of large-scale solar thermal plants used in district heating and industrial processes. FDD simplifies a number of operational and maintenance tasks, thereby reducing costs and enhancing return on investment. Although many approaches have been developed in various industrial fields, there is no definite solution for all cases. Limited work has been conducted in the field of large-scale solar thermal plants, showing that current approaches to fault detection do not cope well with uncertainty and have limited possibilities for diagnosis. The work proposed here will focus on selecting and developing a relevant approach for fault detection and diagnosis of large solar thermal plants. We will especially leverage advanced techniques in sensor data analysis to produce robust algorithms that can cope with uncertainty and help with the automated diagnosis of faults.

This position is open until it is filled.

Département: Département Thermique Biomasse et Hydrogène (LITEN)
Laboratory:
Start Date: 01-10-2015
ECA Code: SL-DRT-15-0904
Contact: cedric.paulus<στο>cea.fr