Quad‐Rotor Modeling and Attitude Control Using State‐Dependent ARX Type Model

Abstract

A state-dependent autoregressive with exogenous variables (SD-ARX) model whose functional coefficients are approximated by sets of radial basis function (RBF) networks is proposed to describe the dynamic behavior of a quad-rotor in this paper. This model is identified offline and used as an internal predictor of a receding horizon predictive controller to address the quad-rotor's attitude control issue. In addition, the physical constraints of the system have been also taken into account during the controller design process. The results of real-time control on a quad-rotor aircraft illustrate satisfactory modeling accuracy in a large operating range and good performance of control approach proposed in this paper.

Advanced Autopilot Systems

“…In the former case, however, fast and accurate on-line estimation of a complicated model providing a good fit to a nonlinear process may be difficult in actual application. RBF-ARX model-based nonlinear MPC algorithms have been investigated both in simulation and in real industrial applications (Peng et al 2004(Peng et al , 2006(Peng et al , 2007(Peng et al , 2009(Peng et al , 2010(Peng et al , 2011Qin et al 2014;Zeng et al 2014;Wu et al 2012), where the satisfactory nonlinear modeling accuracy and significant effectiveness and feasibility of the algorithms have been verified. Furthermore, some stability conclusions on the RBF-ARX model-based nonlinear MPC were also given in Peng et al (2007Peng et al ( , 2011.…”

Nonlinear and Adaptive Intelligent Control Techniques for Quadrotor UAV – A Survey

“…A complete and precise quadrotor mathematical model is difficult to build due to a number of complex aerodynamic forces which act during flight. The more complete mathematical model helps in designing more accurate and robust flight controller [1,[20][21][22][23]. Quadrotor model presented in this study takes an account of gyroscopic effect which occurs due to the rigid body and propeller rotation [8,24].…”

Modeling and Control Approach to Coupled Tanks Liquid Level System Based on Function‐Type Weight RBF‐ARX Model

“…After several adjustments, the final selected parameters of the controllers are those that can achieve the best control performances, which are all well-tuned values to make the liquid levels track the reference signals as well as possible in the middle level experiments. The final parameters of the four controllers are given in Table II, in which R1 and R2 are the diagonal elements of the control weighting matrixes R 1 = diag{[R1]} Nu and R 2 = diag{[R2]} Nu , and Q is the diagonal element of the output error weighting matrixes Q = diag{[Q]} Np in (23).…”

“…O((rN u ) 3 ), where r is the number of control variables and N u is the control horizon. For the water tank system, if their r and N u are the same, the computation complexity of the FWRBF-ARX-MPC is almost the same as that of the RBF-ARX-MPC when solving the QP problem (23). However, the FWRBF-ARX-MPC must compute the function-type weights depending on the working-point state in RBF networks, so the FWRBF-ARX-MPC needs to spend a little more time than the RBF-ARX-MPC to calculate the corresponding matrix in the algorithm.…”