Neuro: adaptive command systems for very maneuverable flying objects

  • Authors:

  • Lungu Mihai;Jula Nicolae;Lungu Romulus;Cepisca Costin

  • Affiliations:

  • Avionics Department, University of Craiova, Faculty of Electrotechnics, Craiova, Dolj, Romania;Military Technical Academy of Bucharest, Romania;Avionics Department, University of Craiova, Faculty of Electrotechnics, Craiova, Dolj, Romania;Electrical Engineering Faculty, University Politehnica of Bucharest, Romania

  • Venue:
  • WSEAS Transactions on Circuits and Systems
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

One presents a dynamic model of the very maneuverable flying objects (A), which expresses the dependence between the vector formed by angles of A regarding aerodynamic trihedron and the vector of angular velocities of A or the vector of linear acceleration components in rapport with the trihedron related to A. Control structures (stabilization of movement) are also presented. These consist of loops after angles, angular velocities and linear accelerations and an adaptive control loop with neuronal network for dynamic inversion errors compensation of the non-linear function which describes unknown system of the dynamic model of A. Adaptive command is projected upon stability theory using Liapunov functions [1], [2], [3]. As calculus examples one presented two systems and stabilization models of the very maneuverable rocket's longitudinal move [4]. Time evolution of the attack angle for the studied cases is also presented.