Fault detection and measurements correction for multiple sensors using a modified autoassociative neural network

Periodic manual calibrations ensure that an instrument will operate correctly for a given period of time, but they do not assure that a faulty instrument will remain calibrated for other periods. In addition, sometimes such calibrations are even unnecessary. In industrial plants, the analysis of signals provided by process monitoring sensors is a difficult task due to the high dimensionality of the data. A strategy for online monitoring and correction of multiple sensors measurements is therefore required. Thus, this work proposes the use of autoassociative neural networks, trained with a modified robust method, in an online monitoring system for fault detection and self-correction of measurements generated by a large number of sensors. Unlike the existing models, the proposed system aims at using only one neural network to reconstruct faulty sensor signals. The model is evaluated with the use of a database containing measurements collected by industrial sensors that monitor and are used in the control of an internal combustion engine installed in a mining truck. Results show that the proposed model is able to map and correct faulty sensor signals and achieve low error rates.