Adaptive Computing Technologies for Diagnostics and Control of UAVs With Fault-Tolerant On-Board Software

Abstract

The article deals with the issue of studying the principles and developing approaches to improving the fault tolerance of UAVs onboard software. The questions of analysis and study of the resource intensity of control and diagnostic algorithms, and the technology of balancing computational tasks using the methodology of distributed computing are formed. The use of hybrid modeling methods is appropriate in solving the problem, since in the work under the control of the UAV we will understand the purposeful impact on the functional processes that describe the implementation of the flight plan, route, selection and the mission as a whole. In this case, the established resource and time constraints must be taken into account. The scientific problem that the project is intended to solve is the creation of new adaptive distributed technologies for diagnosing and managing fault-tolerant on-board software for UAVs based on hybrid modeling methods. In the article, a methodology for modifying the algorithms for monitoring and diagnosing on-board software is formed, which is consistent with the multi-criteria characteristic of possible typical scenarios of UAV behavior.