Predictive Control Of Nonlinear Aeroelastic System Using Neural Network

The aeroelastic studies the question of the structural dynamics of the elastic object.The aeroelastic effect is a phenomenon under the interaction of aerodynamic force,inertia force and elastic body.Because the structure is not absolute rigidity,it has a certain degree of flexibility,when the flow of wind speed to a critical value,the aircraft structure will have a larger deformation,resulting in fatal damage.The aeroelastic system,which uses the two-dimensional wing section as the system model,contains strongly structural nonlinear,multivariate,uncertainty and control surface constraints.Therefore,the control method based on predictive control is used to control the nonlinear aeroelastic system so that the state of the system can reach the steady state quickly.In the predictive control,the accuracy of the prediction model plays an important role in the control effect.This paper presents a novel,fuzzy wavelet neural network(FWNN)structure to identify the nonlinear aeroelastic system.First,the proposed FWNN is developed from the Wavelet Neural Network and interval type-2 fuzzy logic system which has the advantage to approach parameters of aeroelastic system with uncertainty.Additionally,taking the rapidity and accuracy of the identification into account,the system is characterized by a set of fuzzy IF-THEN rules,and the consequent parts of which is designed to be single hidden layer wavelet neural network;And then,the sliding mode algorithm based on Lyapunov stability theory is introduced to obtain parameter update rules,so as to ensure the identification error converge faster under the condition of parameter uncertainty.Finally,numerical simulation for the two-dimensional wing section is investigated to verify the effectiveness of the proposed method.In the rolling optimization step,the paper uses the heuristic Gauss Particle Swarm Optimization(GPSO)algorithm to optimize the constraints.When the position of the particle swarm is initialized,the Gauss operator is introduced to enhance the local searching ability of the particle in the search process.In the subsequent particle position update,the Gaussian function is introduced to the optimization algorithm to gradually reduce the magnitude of the particle position adjustment along with the optimization,so as to ensure global search optimization ability and late approximation effect of early simulation.Finally,the numerical simulation of the two-dimensional wing section with structural nonlinearity is carried out.This paper studies the control effect of the system with parameter perturbation and different control limits of the rudder,and compared with the traditional PID control method and model reference adaptive control method.At last,it is concluded that the predictive control method based on neural network has obvious advantages.