Direct adaptive neural control of a quadrotor unmanned aerial vehicle

Abstract

This paper presents the design of a direct adaptive neural network (DANN)-based feedback linearization (FBL) controller for a multi-input multi output, unstable, nonlinear and underactuated quadrotor UAV. A full system identification was performed on the quadrotor using radial basis function neural networks (RBFNN). The DANN controller was benchmarked against a conventional PD controller which was tuned using Ziegler-Nichols tuning method. The designed controller was found to be capable of accurately tracking the desired output variables simultaneously (altitude, roll, pitch, and yaw angles respectively) and also was robust to parameter variations and disturbance input.

Direct adaptive neural network-based sliding mode control of a high-speed, ultratall building elevator using genetic algorithm

“…Due to the excessive run times recorded during GA-tuning of the initial RBFNN parameters, a different optimisation method would likely be required. Alternative global optimisation techniques that have been suggested for selection of RBFNN-SMC parameters include differential evolution (DE) and particle swarm optimisation (PSO) [23]. Additionally, RBFNN-SMC design utilising the minimum parameter learning (MPL) method could be implemented to reduce the number of adaptive parameters [7], thereby easing computational load and reducing run time during optimisation.…”