Abstract: The invention relates to a diagnostic system which involves the use of a piece of observed information (5) coming from a process (1) monitored by a programmable automaton (10). The diagnostic system includes: an observation module (30) which collects, at input, the information observed (5) and delivers, at output, a piece of processed information (35) to a learning module (40); and a detection module (20) which collects, at input, the observed information (5) in order to apply a detection process (21) thereto including a diagnostic function (45), which comes from the learning module (40), in order to deliver a diagnostic result (25) at output. The observation module (30) and the detection module (20) are executed by the processing unit (11) of the programmable automaton (10) in the form of function units in accordance with standard IEC1131-3. The learning module (40) elaborates the diagnostic function (45) using an algorithm based on neural networks or on fuzzy logic.

Abstract: The invention relates to a diagnostic system which involves the use of a piece of observed information (5) coming from a process (1) monitored by a programmable automaton (10). The diagnostic system comprises: an observation module (30) which collects, at input, the information observed (5) and delivers, at output, a piece of processed information (35) to a learning module (40); and a detection module (20) which collects, at input, the observed information (5) in order to apply a detection process (21) thereto comprising a diagnostic function (45), which comes from the learning module (40), in order to deliver a diagnostic result (25) at output. The observation module (30) and the detection module (20) are executed by the processing unit (11) of the programmable automaton (10) in the form of function units in accordance with standard IEC 1131-3. The learning module (40) elaborates the diagnostic function (45) using an algorithm based on neural networks or on fuzzy logic.