Indoor lighting fault detection and diagnosis using a data fusion approach

In this paper, an innovative and automated fault detection and diagnosis (FDD) approach based on high-level correlation rules in order to improve reliability, safety and efficiency of a supervised building is presented. The proposed method is based on the data fusion of different measurements, using their fuzzification and aggregation through suitable operators, in order to get dimensionless severity indicators able to diagnose faults and to identify the possible causes (ranked according their severity) generating them. Thus, a set of possible anomalies that can occur in a building and the correlation with measured physical quantities were identified. Experimentation of this FDD technique was applied to indoor lighting of a real office building. The proposed method was validated over a onemonth period with the aim of detecting anomalous consumption events, considering when and in which circumstances they occurred. After this stage, the FDD system was performed in real time operation.