Adaptive NN Formation Control for Autonomous Vehicles with Uncertain Dynamics

In this brief, we discuss the formation control of autonomous vehicles. The formation structure uses a leader follower strategy. The vehicles have underactuated characteristics and unknown dynamics. Neural network (NN) is used to approximate the unknown nonlinear function, and the model of uncertain system is established. Meanwhile, the command filtering technique is applied to avoid the problem of burst intricacy intrinsic in the backstepping procedure. Only internal sensors are used to measure and control the line-of-sight range and angle, and no additional instructions about leaders are required; besides, the NN formation controller developed can capture and approximate the unknown dynamics of the vehicle, which reduces the requirement for the accuracy of the vehicle dynamics model. The NN system can also capture and approximate the nonlinear terms of the vehicle, such as wind resistance and friction. The stability of the controller is proved by Lyapunov function. The semi global uniform limit boundedness of the closed loop signal is also guaranteed. Simulations verify the effectiveness of the proposed control strategy.