Research On Adaptive Fault-tolerant Control Technology For HSV With Uncertain Control Direction

Hypersonic vehicles have wide application prospects in military and commercial fields due to the high flight speed.However,the complexity and specificity of the dynamics make the controller design for hypersonic vehicles a challenging problem.This project research focuses mainly on the uncertain control direction problem of hypersonic vehicles,which is caused by the time-varying aerodynamics coefficients,the reverse-type elevon fault and the complex structure of the control system.Based on Nussbaum-type function,adaptive fault-tolerant control strategies are developed for the hypersonic vehicle longitudinal dynamics model in the presence of uncertain control direction,uncertain parameters,unknown elevon faults,input constraints,extra disturbances and nonminimum phase.The designed control system ensures not only the stability of the closed-loop system,but also the output tracking performances.Firstly,the background and significance of this project research are expounded.Then,an overview of the current research status of the longitudinal control for hypersonic vehicles,the uncertain control direction problem and the control design for systems with nonminimum phase are described respectively.Thereafter,two common used plant models,the longitudinal dynamics model for a general hypersonic vehicle(GHSV)and an air-breathing hypersonic vehicle(AHSV),are introduced.Secondly,the adaptive fault-tolerant altitude tracking controller and the adaptive velocity tracking controller are developed respectively for the altitude subsystem and the velocity subsystem of a GHSV.Based on the nominal fault-tolerant controller without considering the uncertainties,an online adaptive estimator is introduced to estimate the uncertain parameters while a Nussbaum gain function is introduced to estimate the uncertain control direction,thereby forming an adaptive fault-tolerant controller.The adaptive laws for the parameter estimates and the Nussbaum function index are selected from the closed-loop tracking error dynamics.Through stability analysis and simulation study,the effectiveness of the proposed design is demonstrated.Thirdly,the Nussbaum gain adaptive control method proposed before is extended to the MIMO longitudinal dynamics model of a GHSV.From the parameterized input-output dynamics,the nominal fault-tolerant controller is designed.Thereafter,in order to deal with the uncertainties,we combine the nominal controller with an online parameter estimator,forming an adaptive fault-tolerant controller.A Nussbaum gain function is introduced in the elevon control channel to estimate the uncertain control direction.Besides,an anti-windup compensation strategy is developed since the commands generated by the designed controller may exceed the admissible input constraints.The asymptotic convergence of the output tracking error and the stability of the closed-loop system are rigidly proved.The theoretical result is substantiated by simulation study.Finally,a Nussbaum-type function based adaptive fault-tolerant controller is designed for the longitudinal dynamics of an AHSV in the presence of nonminimum phase,uncertain control direction,uncertain model parameters and unexpected elevon faults.To begin with,the direct link between the output dynamics and the internal dynamics is built.Using the stable system center method,the ideal internal dynamics is constructed and is used as the reference trajectories of the internal states.Then,a two-layer adaptive controller is developed,where the outer-layer control law inverts the output dynamics into an inner-layer control law while the inner-layer control law is designed using feedback linearization based on an output redefinition.The adaptive laws for the parameter estimates and the Nussbaum function index are chosen appropriately.The designed controller can achieve the output tracking and ensure the stability of the internal dynamics simultaneously.It can be proved that the upper bound of the tracking error is small enough and the whole system is stable.The control design is also substantiated through simulation.