Algorithmic Support for an Adaptive Automatic Monitoring System

Abstract

Modern automatic monitoring systems (AMS) solve the problems of analyzing the flows of measured data coming from the object under study. Solving the problems of regulation and monitoring requires a timely response of the devices to changing parameters of the input signals. Adaptive automatic monitoring systems correspond to these requirements, the construction of which is feasible based on programmable logic circuits, microcontrollers, real-time computers, neural network and expert technologies. At the heart of any of these systems is algorithmic support that describes the principle of action, which determines the operations and the sequence of their execution during the operation of the AMS. One type of adaptation is a parallel-serial analysis of deviations of input values from nominal values, which allows to increase the reliability of monitoring due to parallel calculations of the maximum difference and sequential detection of an allowable deviation.