Adaptive Nonlinear Guidance Law Using Neural Networks Applied to a Quadrotor

The NonLinear Guidance Law (NLGL) is a geometric algorithm commonly employed to solve the path following problem on different unmanned vehicles. NLGL is simple (does no depend on the model of the vehicle), effective and has only one tunning parameter. Its control parameter (L) depends on various factors, such as the velocity of the vehicle, the shape of the reference path and the dynamics of the vehicle. This paper analyses the effect of parameter L on the performance of NLGL when it is applied to a quadrotor vehicle. An Adaptive NLGL, which includes a velocity reduction term, is proposed. Stability proofs are given. Simulation results show that the proposed algorithm enhances the performance of the standard NLGL. Furthermore, it has no parameters to tune.