A neural gust load alleviator for aircraft model using active control

For an aircraft flying in atmosphere a BP neural network controller based on the active control technology is designed. The design goal is to reject the influence of a rotary gust disturbance to the normal overload of the aircraft. In order to improve the dynamic response of such aircraft, the active control technology which will act on the longitudinal control is proposed. The designed controller produces the necessary variation law to improve the ride quality of the passenger. Because the BP net could approximate any nonlinear mathematical model, a kind of BP neural inverse network is presented. However, since the BP algorithm is easy to fall into local optimal value, the particle swarm optimization (PSO) strategy is adopted to train the parameters of the network. Simulation results show that the gust load alleviation (GLA) system designed by neural network could obtain good robust stability, and the capability of restraining gust turbulence as well as measurement noises can be achieved by using the method introduced in this paper.