Sistema en lazo cerrado para el diagnóstico de fallos de sistemas de eventos discretos utilizando redes de Petri interpretadas.

This paper presents the implementation of an online fault diagnostics, for the detection of operational faults present in the pilot water processing plant, within the framework of the REAGRITECH project of the UNESCO chair of sustainability, modeled as a system of discrete events (SED), making use of interpreted Petri nets (IPN). An analysis of the operation of the real system is carried out, from which some operating scenes are identified, which allow the construction of a controller in charge of representing, in a simulated way, the real behavior of the plant. From this, an input-output (I/O) information matrix is obtained, where the inputs correspond to the control command signals, and the outputs are the sensor signals. These data are entered as parameters to an identification algorithm, which delivers as a result the model of the IPN identified from the data, that is, information of an IPN: incidence matrix A, input function (Fe) with the labels associated with the transitions, the output function (Fs) and an output matrix φ that relates the sensors to each slot in the system. Based on the identified model, a detectability analysis is carried out to find out if the system is detectable or not by events, building a diagnostic from predefined faults. It is concluded that the diagnostic tool implemented is capable of detecting faults present in the pilot water processing plant, within the framework of the REAGRITECH project, making an analysis of the marking in the diagnostic places, and in the Post risk places of the system. [ABSTRACT FROM AUTHOR]