ПРОГНОЗУВАННЯ МІЖЕЛЕМЕНТНОЇ ВЗАЄМОДІЇ В СИСТЕМАХ АВТОМАТИЧНОГО КЕРУВАННЯ АВІАЦІЙНИХ ДВИГУНІВ

Abstract

The article is devoted to the development of modern models and algorithms for information processing for the prediction, diagnosing and maintenance tasks of inter-element interaction in automatic control systems of aviation gas turbine engines. Considered the model of inter-node interaction with the definition of power active points and their neural structures with visualization in the form of scalogramm. Presented a graph of changes the power node characteristics of the automatic control systems of gas turbine engine and the distribution histograms of practically normal states and faults with defect properties on the scale of confidence coefficients based on the control algorithm the prediction and diagnostics processes according to the data of the system survey.The decision support system during the research process has two stages of functioning:1) at the stage of prediction and diagnostics, the measurement of the power characteristics of the active interaction points is determined, and corresponding elements of the models are taken into account the power states formed on the nominal energy state of the nodes;2) at the stage of maintenance and repair, while observing the painted nodes, determine the type, intensity and duration of influence in order to normalize the power of the nodal active points of interaction (to bring the state to a dark brown colour). At the same time, the dynamics of the correction process is well traced, visually visible to which accompanying situations is directed the impact and how it will affect the functioning of the linked nodes and /or systems.During the developing an appropriate system of inter-element interaction in automatic control systems of gas turbine engines, it is possible to receive an educational and control sample to support decision-making, check the reliability of the operation of the received rules, and if necessary, a training procedure can be implemented to improve the accuracy of the classification by training and control samples.The developed rational technology of building the architecture of nodes in the automatic control systems of aviation engines will create high-performance systems of a new generation with a flexible, easily variable structure, increase reliability and durability of operation of automatic control systems of aviation engines.