Aerospace Instrument-Making Annotation << Back ONBOARD NEURAL NETWORK OF AIRCRAFT CONTROL SYSTEM BASED ON CONTINUOUS ANALOG PROCESSORS N.L. Dembitskiy Experts in the field of aviation distinguish a combination of speed, maneuverability and controllability as one of the main problems in the development and improvement of the tactical and technical indicators of an aircraft. Automatic control systems to the greatest extent determine the flight characteristics of the aircraft. For a maneuverable aircraft, one of the main indicators of the effectiveness of automatic control systems is the reaction rate.For example, when performing intense flight tasks of a fighter plane, delays of 0.1 second make high levels of aerobatic quality unattainable, absorbing most of the phase margin in the control loop. Digital systems do not always provide the necessary speed for generating control signals. Significant quantization errors and program crashes can play a significant role in reducing the effectiveness of digital control. These problems create the prerequisites for finding alternative ways of developing automatic control systems for aircraft. In this paper, we consider an approach based on the use of continuous analog processors, which combine the logical and computational functions of signal processing. Based on analog processors, a functional-logical neural network is created, which is built into the aircraft aggregate control system. Key words: highly maneuverable aircraft, aircraft, automatic control systems, continuous analog processors, neural networks. DOI: 10.25791/aviakosmos.11.2019.988 Contacts: - Pp. 13-20.