IJCAS

In this research article, a novel distributed neuroadaptive sliding mode control (SMC) algorithm has been developed for three-dimensional (3-D) formation control of a swarm of unmanned aerial vehicles (UAVs). The UAVs (all identical) track a virtual leader. A sliding mode and a novel robust controller, based on classical SMC, have been conceptualised and developed, which shape the formation and suppress disturbances effectively. Radial basis function neural networks (RBFNNs) have been employed for the function approximation of external disturbances and the generation of the robust control signal. An adaptive law has been proposed for tuning the weights of the neural networks online. The unique feature of the developed algorithm has been its mathematical simplicity and low complexity in implementation as compared to current algorithms. Furthermore, the finite time for formation (i.e., the error convergence time) can be controlled by varying the robust controller gain. A low-pass filter has been implemented, which clears the chattering and oscillations from the control signal. Both time-varying and time-invariant formations have been achieved. The Lyapunov stability analysis proves the overall stability. Numerical simulations show the validity of the proposed algorithm. The proposed algorithm has been tested in a Gazebo simulation environment for the time-invariant case. The Gazebo simulation results confirm the validity of the proposed algorithm. A comparison of control performance indices with a recent work of similar nature yields superior results

Keywords: Formation control, Lyapunov stability, RBFNNs, sliding mode control, UAVs.