Research On Self-Healing Control For Hypersonic Flight Vehicles With Composite Sensor Faults

The flight environment of the hypersonic flight vehicle has the severe characteristics of high speed,low pressure,high temperature,strong radiation,electromagnetic interference,lightning and strong convection.The components of the vehicle often have external mechanical vibration,pipeline dropout and electronic device failure,which may lead to sensor system suffer from composite faults and have serious consequences.Research on self-healing control of hypersonic flight vehicles under sensor composite faults thus has become a significant topic.In this paper,we take the longitudinal model and the attitude angle model of the hypersonic fight vehicle as the research object.The fault detection and estimation problems of various composite faults are studied,and the corresponding self-healing control methods are designed.The main contents of the paper are as follows:(1)A self-healing control method based on adaptive and backstepping sliding mode fault estimation observers is designed for the multi-sensor faults in the height and velocity channel of the hypersonic flight vehicle longitudinal model.Firstly,the longitudinal affine nonlinear model of hypersonic vehicle is established by feedback linearization method,and the nominal sliding mode controller is designed to make the output track the given reference command.A nonlinear adaptive observer method is proposed to estimate the fault of the velocity channel.By using the system parameters affected by the coupling as the estimation term,the backstepping sliding mode observer is designed to accurately estimate the height channel fault.The sliding mode controller is compensated by the fault estimation result to ensure the system re-track the given reference command.(2)Taking the more complicated time-varying sensor cascading faults and external interference into consideration,a fault diagnosis and fault-tolerant control scheme based on multi-dimensional generalized observer is designed.On the base of the cruise phase longitudinal T-S fuzzy model,the external disturbances and nonlinear terms are considered to enhance the representativeness of the model.For the sensor cascading failure,the fault detection observer is firstly established to detect the fault.Then the multi-dimensional generalized observer method is designed to solve the problem that the normal generalized observer does not ideally estimate the cascading failure due to the coupling effect.According to the obtained fault estimation results,a robust fault-tolerant controller capable of handling system nonlinearity and sensor failure is designed to stabilize the system.The designed observer and controller satisfy the H? performance index and are robust to interference.(3)For the sensor composite faults occurring in the hypersonic reentry vehicle attitude angle system,a self-healing method based on indirect adaptive sliding mode and dynamic surface technology is designed.Firstly,for the attitude angle system with sensor composite faults and interference,a nonlinear fault detection observer is designed to generate detection residuals to detect faults.An indirect adaptive sliding mode virtual controller and dynamic surface controller design method with fault-tolerant items are proposed.Sensor composite faults are quickly compensated during controller design,eliminating the need for additional fault estimation observer designs,ensuring that faulty systems are recovered with good tracking of reference instructions while reducing system complexity.The above-mentioned fault detection,fault estimation and self-healing control methods were compared based on Matlab/simulink.The rapid prototyping experiments were carried out on some results using the Links-Box platform.The results show that the proposed scheme can quickly detect faults and accurately obtain fault estimation information while effectively compensating them to ensure system stability and tracking performance.