Research On Nonlinear Modeling And Control Of Small Unmanned Helicopter

The dynamic model of a small unmanned helicopter has the characteristics of non-linear,underactuated and non minimum phase.It has high control sensitivity and poor anti-interference ability.Its autonomous flight is a challenging control problem.The control accuracy and robustness of classical and linear flight control systems still have much room for improvement.Based on the small unmanned helicopter system platform constructed in the laboratory,this paper uses the nonlinear modeling and nonlinear parameter identification method to establish the mathematical model,and uses the adaptive robust nonlinear control method to design the controller,aiming to obtain a more accurate and robust flight control system.The strict analysis and comprehensive experimental test are combined to ensure the effectiveness of theoretical design in practical engineering application.The main contents and innovations are as follows:(i)A new identification method based on the power model of the main rotor is proposed,which only needs to apply the active excitation signal in the vertical channel of the small unmanned helicopter,so that the yaw dynamic model can be identified on the basis of the vertical dynamic model,reducing the parameters to be identified,simplifying the identification flight experiment,and improving the identification efficiency and safety.An adaptive differential evolution identification algorithm is proposed.Based on the traditional differential evolution algorithm,the global search ability is further improved,the local optimal trap is effectively avoided,and the optimal solution can be found quickly.Compared with the real system,the identified nonlinear model has high fidelity.(ii)The traditional adaptive RBF neural network control method has a large amount of computation,and the number of adaptive laws increases exponentially with the increase of the number of hidden layer nodes.An adaptive RBF neural network control method with minimum learning parameters is proposed.It only needs to design an adaptive law for the square of the Euclidean norm of the ideal weight matrix,which reduces the calculation burden of the airborne system,and takes into account the robustness of the control system and the engineering realizability.A simple and reversible expression is proposed to associate the non physical input with the physical input,which breaks through the traditional attitude distributed control structure,making it possible to design an integrated three-dimensional attitude control system for an unmanned helicopter.The effectiveness and robustness of the method are verified by simulation and real experiment.The autonomous flight control of the small unmanned helicopter with strong robustness and large envelope is realized.(iii)An adaptive quasi optimal high order continuous sliding mode control method is proposed,which can specify the convergence speed.The dynamic response of small unmanned helicopter is fast,and the control system is required to be fast.The model of the system is transformed into the integral chain form,and the finite time convergence quasi optimal control is designed for the integral chain system.This method can converge quickly and specify the convergence speed without knowing the bounds of system uncertainty.The adaptive law solves the problem of over estimation of sliding mode gain and has the advantages of suppressing chattering.The speediness index,bandwidth and phase delay index of attitude control system all reach the first level flight quality of ADS-33E-PRF.The effectiveness and robustness of the whole control system are verified by simulation and real flight experiments.