Intelligent Support of Multilevel Functional Stability of Control and Navigation Systems

Abstract

Relevance. Ensuring functionally steady operation of dynamic objects' movement in the presence of abnormal situations is one of the most prospective research problems in the area of technical diagnostics and restoration of system operability. Use of intelligent control methods and algorithms for solving this scientific and technical task can significantly extend the functionality and improve the performance of control systems. The aim is the formation of an approach to determine the control effect that ensures the functional stability of the CNS in the presence of abnormal situations. Method. The concept of ensuring functionally stable control of dynamic objects' movement has been offered. Well-known diagnostic methods and tools have been systematized and on that basis new models and methods for deep diagnosis of the functional state of the control and navigation systems up to the reason of abnormality have been developed. Models and methods for multi-level parrying of the reason of abnormal situations through control over the diagnosis have been synthesized, they use such redundant resources as signal and parametric adjustment, reconfiguration of algorithms and commutation of equipment. Results. The results of the solution of a number of combined scientific and technical problems aimed at the multilevel ensuring of the functional stability of CNS have been presented. Conclusion. Analyses of tendencies of theoretical investigation and practical achievements in ensuring functionally steady control of CNS at occurrence of abnormal situations has been carried out. The concept of intelligent support of the functionally steady control of CNS has been developed. The concept is based on the principles of multilevel hierarchical diagnosis of CNS to the parried reason of abnormal situation, as well as on the situational approach to eliminate the consequences of failures in accordance with the level of its application. As a result of analysis and synthesis of known approaches to diagnosing the technical state of dynamic objects, the new models and methods for deep diagnosis of the CNS' functional state have been formed, that makes possible to determine the failure up to the reason of abnormality. Models and methods for multi-level parrying of abnormality through control over the diagnosis with the use of available intelligent on-board resources have been offered. A hardware-software complex for experimental research of theoretical positions have been developed, and conducted experiments have been verified the possibility of extending the functionality of the system to counter abnormal situations. As a result of conducted experiments in real-time scale, diagnosis and restoration of CNS' functional state during a time not exceeding 42.2%-48.4% of the transient time of the system in normal operation mode have been implemented.