Abstract: A method for detecting anomalies in a digitized complex signal analyzed by a detection unit, including a machine learning and a diagnosis of the intensity and/or the rarity of an anomaly, the learning including the steps of: 1.1 selecting sequences of values of the signal; 1.2 transforming the signal to extract therefrom characteristics of a type easily extracted by a human eye; and 1.3 reducing number n of digital data by an automatic compression; the diagnosis including the steps of: 2.1 applying steps 1.1 to 1.3 to a polling window (Fk) likely to include an anomaly; 2.2 comparing the obtained vector with a reference defined according to the same transformation and compression structure.

Abstract: A method for monitoring a system based on a set of k performance indicators, Xj(t), each of which is defined at successive times t, j being an integer varying between 1 and k. This method includes, for each indicator Xj, the steps of performing an observation of a sequence of s values, s being an integer, of the indicator and reordering this sequence into a reference list ordered by increasing values; and determining the relative rank, REL[Xj(t)], in the reference list of any new value XJ(t) of the indicator, this relative rank being equal to the rank of the new value divided by number s. The present invention also provides the computation of scores based on these relative ranks. Essential indicators and merged scores are deduced from these scores.