Resumen
The present invention enables identifying each measured value of the input variables (voltage, current and phase shift angle) of the status of the system (1). Based on the time evolution of the statuses of the system (1), it enables identifying sequences or groups of statuses that describe aggregated behaviours of said system. In particular, it comprises the steps of: training artificial intelligence models; capturing analogue voltage, current and phase shift angle signals (2); synchronously predicting the discrete status (5) of the system (1) at each moment; sequencing and segmenting the discrete statuses (5) of the system (1), obtaining a set of subsequences (6); and obtaining a set of behaviour trends of the system (1) based on the subsequences (6), their duration and repeatability.