Adaptive neural network-based active disturbance rejection flight control of an unmanned helicopter

Abstract

To enable the unmanned helicopter to fly autonomously in precise paths and reduce the influence of internal and external unknown disturbances of the unmanned helicopter, this paper proposes the adaptive radial basis function (RBF) neural network-based active disturbance rejection controller (ADRC). This controller is abbreviated as RBF-ADRC. Firstly, this paper introduces the flight dynamics model of the unmanned helicopter and the control features of the traditional ADRC. Subsequently, this paper uses modern control theory to establish a state observer and uses adaptive RBF neural network to estimate the unknown total disturbance. Finally, this paper constructs the unmanned helicopter's trajectory tracking control system based on the RBF-ADRC controller. The simulation results of the spiral ascent and the “8”-figure climb maneuver flight prove that the anti-disturbance, robustness and tracking accuracy of the RBF-ADRC are better than the traditional ADRC and proportion-integration-differentiation (PID) control methods.