Fuzzy Sampled-Data Extended Dissipative Tracking for Quadrotor UAVs under Cyber Attack

Although the design of robust controllers using linear control laws and linear matrix inequalities (LMIs) is highly useful, the strong nonlinearity of quadrotors makes this challenging. Some studies have addressed this issue using Takagi-Sugeno (T-S) fuzzy modeling; however, the results are still very limited. This study addresses the sampled-data control problem of quadrotors under cyber attacks based on linear control laws. Unlike previous studies, a dissipative tracking controller design method that includes some different performance indices is proposed. The two highly coupled and nonlinear subsystems of the quadrotor (altitude and attitude) are expressed by the T-S fuzzy model, which is appropriate for designing linear controllers. The problem of the sampled data control is addressed by employing an input delay approach. Attackers can inject malicious information into the control signals, thereby degrading system performance. In this context, less conservative sufficient conditions for achieving stability and extended dissipative tracking control performance of the closed-loop system are provided within the framework of the Lyapunov theory. Experimental results demonstrate that the proposed method achieves significantly improved tracking performance compared with recent nonlinear controllers, namely neural adaptive sliding-mode control (SMC) and backstepping radial basis fuction (RBF) control. Specifically, the proposed controller reduces the RMS tracking errors by approximately 13.8%, 27.4%, and 3.5% relative to neural adaptive SMC, and by 2.1%, 1.8%, and 2.7% relative to backstepping RBF, along the x, y, and z axes, respectively. Furthermore, the proposed method maintains stable tracking even under strong wind disturbances (5-7 m/s) and during cyber attack intervals, thereby confirming its robustness and effectiveness.