Naučno-tehničeskij Vestnik Informacionnyh Tehnologij, Mehaniki i Optiki (Nov 2015)

СONTROL FOR QUADROCOPTER WITH COMPENSATION OF WIND DISTURBANCE

  • I. V. Petranevsky,
  • O. I. Borisov,
  • V. S. Gromov,
  • A. A. Pyrkin

DOI
https://doi.org/10.17586/2226-1494-2015-15-6-1045-1053
Journal volume & issue
Vol. 15, no. 6
pp. 1045 – 1053

Abstract

Read online

Subject of Research. The paper presents an output control approach for a class of nonlinear MIMO (Multiple Input Multiple Output) systems. Quadrocopter is chosen to illustrate proposed adaptive control approach based on the high-gain principle, the so-called, “consecutive compensator”. Basic Concepts. The dynamic properties of quadrocopter can be described by a non-linear multi-channel system. Accordingly, the first step is accomplished by its decomposition, whereby the model is divided into two parts: static conversion and several dynamical channels with one input and one output). After the decomposition, a derivation of the simplified mathematical model of the quadrocopter is carried out; and, in addition to the absolute coordinate system, the local system rigidly connected with the robot is introduced. The appropriate coordinate transformation is taken into account. Interim Results. Decomposition of the mathematical model gives the possibility to create control system in two steps. Firstly, the law is applied; in accordance with it the so-called virtual control actions for each model channel of the quadrocopter are generated. A well-known method of robust control, "consecutive compensator", is used here, successfully well-accepted as the most easily implemented one from an engineering point of view. This method gives the possibility to control an object output under its uncertainty quantification. Secondly, the inverse transformation is performed, resulting in real control actions distributed between the system actuators. We assume that wind disturbance affects each channel of the object. What is more, the values and directions of these disturbances are changed negligibly slow. Thus, we can treat them as unknown constants, subject to compensation. Main Result. We have developed a multi-channel control system for the quadrocopter by different methods (decomposition of the mathematical model and the output of its simplified analogue, coordinate transformation, "consecutive compensator"). The obtained closed-loop system has been successfully simulated and corresponding results have been received.

Keywords