Self-monitoring of technical condition in electronic devices with structural redundancy

Abstract

The introduction of hardware redundancy in most cases is used as a method of increasing a reliability of non-repairable (non-recoverable) equipment. In this case, not only the improvement of such an indicator as the probability of failure-free operation can be achieved, but an increase in MTBF as well. The article shows that in the case of introducing redundancy for the restored equipment, additional opportunities arise for organizing self-diagnosis (self-monitoring) of the technical condition of the equipment. On the example of the structure of the popular NonStop solution derived by the Tandem company, specific proposals for the development of real-time signals for indicating the current technical state of the structure in the event of pre-emergency or emergency situations are considered. It is shown, that aforementioned possibilities are practically absent in the majority structures in the case of triple redundancy and significantly improve upon the transition to structures of higher redundancy. Another important and useful feature of such structures is the ability to adapt the signal generation algorithm when a part of the equipment fails in the structure. In the case of "Tandem"-type structure after the failure of one of the elements, the implementation of the adaptation algorithm leads to a structure with triple redundancy of the majority type, which has significant advantages over double-redundant. The analysis of "Tandem"-type structures for various starting positions was carried out. Analytical expressions are obtained for the average transition times from the "Tandem" scheme to the triple redundancy scheme (hereinafter, to the double redundancy scheme with subsequent absence of redundancy), the average reverse transition times in the presence of recovery possibilities, and, based on them, coefficients of readiness are calculated.