Research On Aerodynamic Parameters Online Identification And Longitudinal Control For Hypersonic Vehicle

Hypersonic vehicles are widely used in a variety of maneuvering tasks due to their good maneuverability.However,due to the strong nonlinearity and uncertainty of hypersonic vehicles,the inconsistency of the aerodynamic parameters of the heavens and the earth is particularly serious,which puts high requirements on the accuracy of the aircraft guidance and control system.Therefore,if the more accurate pneumatic parameters can be obtained through online identification,it will relieve the pressure on the aircraft control system.In this paper,the problems of on-line identification of aerodynamic parameters of hypersonic vehicles and tracking control of cruise segments are studied.According to the real aerodynamic model in the actual flight environment,a linear aerodynamic model is obtained by reasonable linearization,which lays a foundation for subsequent identification.At the same time,according to the motion model of the hypersonic vehicle in the spherical earth,the dynamic equation of the longitudinal plane is established and divided into a height subsystem and a speed subsystem,which are respectively controlled by the attitude controller.In parameter identification,a metric parameter is studied to measure the identifiability of input excitation.The sinusoidal excitations with different amplitudes,different action times and different periods are used respectively,and the parameters of the aerodynamic model of the aircraft are identified online by using the recursive least squares method.The results of the three sets of contrast experiments not only show the impact of different parameters,but also verifies the reliability of the metric parameters proposed in this paper,which can be used to quickly measure the identifiability of the stimulus and facilitate the design of the input.Aiming at the problem that the traditional extended Kalman filter has low precision in parameter identification,this paper combines it with the iterative filtering theory to obtain an iterative extended Kalman filter method.Based on the extended Kalman filter,the latest states estimated are iteratively updated to obtain more accurate states at the current time,thereby improving the accuracy of the algorithm.In the simulation experiment,it is found that the accuracy of the algorithm is improved as the number of iterations increases,but in the actual situation,the real-time requirements should be considered to determine the number of iterations.Finally,based on the aerodynamic model obtained by parameter online identification,an adaptive back-stepping method is used to design an adaptive controller for the two subsystems of the longitudinal plane of the hypersonic vehicle,and a low-pass filter is introduced in the height controller to avoid the problem which called "differential explosion".The results of simulation show that the controller designed can quickly and stably control the aircraft to track the desired signal due to the more accurate aerodynamic model obtained through online identification,while the aerodynamic model with deviation will have steady-state error in the control process.It verifies the effect of parameter online identification on improving control accuracy.