Method of monitoring of a system of operation using empirical and distributions of the standards of cumulative distribution

摘要

Sensors associated with various system parameters provide values of corresponding parameters which are used to construct an empirical distribution process, by for example a firing methodology using a 5 dimensional Markov chain model. In normal use the sensors are periodically read and in response to events and any anomaly is determined by comparing current data with the empirical distribution process. Deviations from normal system behavior provide a quantitative measure of the fault or system anomaly . The elimination of data from one or more sensors in each iteration process identifies one or more sensors linked to the fault or anomaly. Independent Claims are included for details of the method employed