The Digital Technology of Forecasting Failures by Considering Noise as a Data Carrier

In the past, the errors in forecasting such objects as oil-chemical complexes, deep-water stationary sea platforms and communications, hydraulic works, etc. were assumed to be connected with meteorological and assurance characteristics of the element base of information measurement systems. Now they have been improved, but the probability of a failure has remained the same. Analysis shows that the main reason for an inadequate decision from a diagnostic system is connected with the impossibility of detecting the initial state of arising defects with the known methods of analyzing noisy signals [14, 15, 17, 56].

In the literature, a “fault” is defined as the inability of a system to realize required functions [55–57]. The fault is the initial state of the failure, which means the inability of work on a substantial scale. In this case, the fault of certain elements of a system leads to faults of general parts of an object and finally to complete destruction. In this case, the future use of an object is impossible or requires major repairs [55–57].