Scientific Reports (Oct 2022)

Real-time robust generalized dynamic inversion based optimization control for coupled twin rotor MIMO system

  • Nadir Abbas,
  • Xuejun Pan,
  • Abdur Raheem,
  • Rabia Shakoor,
  • Zeeshan Ahmad Arfeen,
  • Muhammad Rashid,
  • Farhana Umer,
  • Nouman Safdar,
  • Xiaodong Liu

DOI
https://doi.org/10.1038/s41598-022-21357-3
Journal volume & issue
Vol. 12, no. 1
pp. 1 – 17

Abstract

Read online

Abstract This work is used to design a novel robust optimization control law augmented with Robust Generalized Dynamic Inversion (RGDI) for continuous varying perturbations in the Twin Rotor MIMO System (TRMS). The perturbations like coupling effect, un-known states, gyroscopic disturbance torque, parametric uncertainties and parametric disturbances are considered as unwanted signal which should be optimized by an efficient controller. The variable structured systems like the TRMS (prototype) have great focus due to its high computational cost with a higher order non-linear behavior. The RGDI based controller designed to remove nonlinear dynamics as well as to avoid singularity issue with the augmentation of stability based mathematical operations (lyapunov stability analysis, controllability and observability matrices ) in the presence of considered perturbations during implementation. In this paper, we develop estimation of state deviation calculation between control angles and desired angles known as Euclidean error norm. The next step was to design RGDI based controller [Sliding Mode Control (SMC) and $${H_\infty }$$ H ∞ optimization] to minimize considered perturbations as well as the computational cost. The sharp (rapid) chattering phenomena in RGDI based SMC reduce the actuators performance that goes towards the failure of actuators. While the RGDI based $${H_\infty }$$ H ∞ optimization overcome the computational cost and minimizes $${H_\infty }$$ H ∞ norm that’s guaranteeing the robust stability as well as robust performance. The robustness of the optimization control technique validated by taking its worst case via MATLAB-Simulation. A real-time implementation applied to evaluate the worth of novel dynamic approach.