10.4122/1.1000001158
Hamioud, F.
F.
Hamioud
Joannis, C.
C.
Joannis
Ragot, J.
J.
Ragot
Fault diagnosis for validation of hydrometric data collected from sewer networks
DTU Library, Technical University of Denmark (DTU)
2005
Conference full text
fault detection
fuzzy logic
linear model
residual
sensor
Laboratoire Central Des Ponts Et Chausse\303\251s
Institut National Polytechnique De Lorraine, Centre De Recherche En Automatique De Nancy
Centre De Recherche En Automatique De Nancy
Institut National Polytechnique De Lorraine
2005
en
10.4122/1.1000001157
application/pdf
The measured values provided by sewerage monitoring systems require a processing for their validation. This paper applies a method of diagnosis on a set of measurements collected from the sewer network of Nantes (France), This method relies on the overall information redundancy, which is captured in a set of linear models. These models generate residuals, which are tested for the detection and localisation of potential faulty sensors. The diagnosis uses the theoretical signature matrix, which is derived from the structure of the models, and qualitatively describes the effect of a faulty measured variable on the residual of each model. Then, fuzzy logic is used to compare the theoretical signature matrix and the actual values of observed residuals, thus enabling the identification of faulty sensors. An example is provided to illustrate this method.