Real-Time Implementation of Neuro Adaptive Observer-Based Robust Backstepping Controller for Twin Rotor Control System

  • PDF / 1,269,709 Bytes
  • 14 Pages / 595.276 x 790.866 pts Page_size
  • 75 Downloads / 207 Views

DOWNLOAD

REPORT


Real-Time Implementation of Neuro Adaptive Observer-Based Robust Backstepping Controller for Twin Rotor Control System Bhanu Pratap · Shubhi Purwar

Received: 11 August 2013 / Revised: 20 November 2013 / Accepted: 12 December 2013 / Published online: 31 December 2013 © Brazilian Society for Automatics–SBA 2013

Abstract In this paper, a robust backstepping controller based on the neuro adaptive observer for the twin rotor multiple-input-multiple-output (MIMO) system is designed and implemented in real time. The twin rotor MIMO system (TRMS) belongs to a class of nonlinear uncertain system having unstable, coupled dynamics. Nonlinearities of the TRMS are estimated using Chebyshev neural network. A tuning scheme based on Lyapunov theory of stability is developed which can guarantee the boundedness of tracking error and weights of the neural network. The proposed observer-based control guarantees the stability of the closed-loop adaptive system and the tracking errors converge to small residual sets in the presence of constraints on the control input. The effectiveness of the proposed observer-based robust controller is illustrated through simulation and experimental results. The real time implementation has been carried out on the realtime TRMS using MATLAB real-time tool box and Advantech PCI1711 card. Keywords Backstepping technique · Chebyshev neural network · Nonlinear coupled systems · Observerbased controller · Twin rotor MIMO system

B. Pratap (B) Department of Electrical Engineering, National Institute of Technology, Kurukshetra, India e-mail: [email protected] S. Purwar Department of Electrical Engineering, M. N. National Institute of Technology, Allahabad, India e-mail: [email protected]

1 Introduction In the past decade, control design of nonlinear systems has attracted an ever increasing interest. There have been significant research efforts on intelligent control (Ge and Zhang 2004; Ge et al. 1999; He et al. 1998), sliding mode control (Elmali and Olgac 1992; Byungkook and Woonchul 1998), robust adaptive control (Yao and Tomizuka 2001; Haddad et al. 2003; Lee and Lee 2004; Kwan and Lewis 2000), and backstepping control (Kwan and Lewis 2000; Zhang et al. 2000; Gong and Yao 2001; Huang and Chen 2004; Wang and Huang 2005). To enhance the control performance of unknown/uncertain nonlinear systems, different kinds of techniques can be integrated, utilizing respective advantages in the control system design. The modeling and control of the TRMS (2006) has gained a lot of attention because the dynamics of the TRMS and a helicopter are similar in certain aspects (Khan and Iqbal 2003, 2004; Kim et al. 2006). Due to unstable, nonlinear dynamics and high coupling effect between two propellers, the control problem of the TRMS has been considered as a challenging research topic. In Wen and Lu (2008), a decoupling control of TRMS using robust deadbeat control technique is designed. The system is decoupled into two SISO systems, and the cross couplings are considered as disturbances. A robust deadbeat control sc