Development of Centralized Multivariable Cascade Control Strategy for Quadcopter System

Abstract

In this article proposes a novel centralized cascade control framework for quadcopter attitude and trajectory tracking that explicitly suppresses secondary cross-coupling effects commonly observed in multivariable UAV dynamics. The nonlinear quadcopter model is transformed into a set of region-dependent linear MIMO state-space representations using a piecewise affine (PWA) formulation, and each region model is converted into a MIMO transfer-function form to support systematic multivariable controller synthesis. To avoid ad hoc loop selection, a hybrid loop-pairing method is introduced to identify feasible pairings and reduce unfavourable inner-loop interactions across operating regions. Based on this pairing, an IMC-based switching enhanced cascade PID/PI design is developed, employing an enhanced PID controller in the outer loop and an enhanced PI controller in the inner loop, with regionwise scheduling to maintain performance over the operating envelope. The proposed method (Ce) achieves the lowest ISE in both Simulation-1 and Simulation-2 for roll \(\left( \phi \right)\) and pitch \(\left( \theta \right)\), showing consistent improvement over both De and Ge. The ISE reductions are 14.34 and 11.29% over De in Simulation-1, and 12.91 and 16.83% over De in Simulation-2 for roll and pitch, respectively, with even larger improvements over Ge. Extensive simulations demonstrate improved tracking accuracy with reduced coupling effects while satisfying state and input constraints. Comparisons with a decentralized cascade controller and a conventional cascade controller further confirm the effectiveness and efficiency of the proposed approach.