Research On Multi-mode Adaptive Control Method For Reusable Launch Vehicle

During the large flight envelop,the reusable launch vehicle(RLV)faces with such dramatic changes of the aerodynamic environments,working modes,actuators,and even the structural shapes from the different flight stages.On one hand,it makes the spacecraft have the characteristics of serious nonlinearity,strong coupling,fast time-varying,and uncertainty;on the other hand,the single control model and control method are increasingly unable to meet the task requirements of RLV’s control system.As a control method quite suitable for such complex systems,the multi-mode control can be combined with other modeling methods and control methods to build a sub-model set,sub-controller set and an adaptive switching strategy for the controlled object,so as to ensure its control system’s strong robustness and adaptability.Considering RLV’s characteristics of working modes,system uncertainties,strong nonlinearity,redundant heterogeneous actuators and so on,a multi-mode adaptive control system is designed in this paper.Meanwhile,it focuses on the study of the sub-model set constructing for different flight phases,the nonlinear adaptive control based on the disturbance compensation,the dynamic adaptive control allocation based on the multi-object optimization,the adaptive switching control based on the control performance assessment.Then,regarding the controllers,control allocation strategies and switching strategies proposed or designed in this dissertation,the respective feasibility analysis,comparison of different methods,and the multi-mode adaptive control system for RLV’s whole flight are verified and demonstrated by the simulation.In view of the large envelop and multiple modes,RLV’s 6-DOF dynamic model is built for the whole flight.According to this 6-DOF model,the corresponding linear control model based on the small disturbance hypothesis,and nonlinear control model without the hypothesis are derived.After the flight timing analysis,by taking the dynamic characteristics,control variables and the actuators of different flight stages into accountant,the feature points are selected,the corresponding neighborhood regions are divided,and then the sub-model set is built,which is the basis of the controller set,control allocation set and switching strategy designed below.For the nonlinear system with parameter uncertainties and external unknown disturbances,a nonlinear adaptive control method based on the disturbance compensation is studied.First,a nonlinear PID(NLPID)controller is designed for the linear control model.Second,for the nonlinear control model with parameter uncertainties and unknown disturbances,an adaptive backstepping(ABS)control method is proposed.Since ABS suffers the problem of “explosion of terms”,an extended state observer based on the Sigmoid function(SESO)is introduced into the control structure to replace the ABS controller’s self-adaptation law.Finally,by combining the advantages of the above methods,a nonlinear adaptive control method is proposed,and the stability of the closed-loop control system is theoretically analyzed.Simulation results verify the effectiveness of NLPID,ABS,SESO and the proposed adaptive control method.In addition,the precision,robustness,disturbance resistance and other control performances of the aforementioned controllers are compared.The dynamic control allocation method based on the multi-object optimization is studied for the redundant heterogeneous actuators.Firstly,with the object function of the allocation error minimization,the control allocation methods’ research is carried out for the same kind of actuators and the different kinds.Secondly,in order to find the integrated optimal solution to the allocation error minimization,control efficiency maximum,fuel consumption minimization and so on,the allocation problem based on the constructed multi-object optimization functions is transformed into the single-object optimization problem by a weighted combination with the dynamic weight coefficients designed by the analytic hierarchy process method.Then,after deducing the theorems of the allocation addition making no difference to the stability of the closed-loop control system,the simulation verification for the control allocation schemes is carried out.Meanwhile,the switching timing of the actuators,the advantages of the multi-object optimization,and the comparison of different control allocation schemes for the heterogeneous actuators are emphatically analyzed.The results not only show the feasibility of the designed allocation schemes,but also demonstrate the better comprehensive performance of the multi-object optimization allocation scheme,by considering the allocation error,control error,fuel consumption,and so on.Aiming at the designed sub-model set,sub-controller set,sub-control allocation set above,an adaptive switching control strategy based on the control performance assessment(CPA)is studied.Firstly,the CPA system based on the analytic hierarchy process method is built as the basis.A CPA function of the process indexes consisting of the error criterion established by the sub-model set and the actual RLV model,is designed as the adaptive switching strategy.Meanwhile,the stability of the multi-mode control system with the designed switching strategy is proved based on the method of mean dwell time.Secondly,taking the CPA function as the optimization objective,a feedback adaptive law based on the steepest descent method is proposed,to optimize the control parameters while ensuring the system’s stability.Finally,on the premise of building the platform of RLV’s multi-mode adaptive control system,the feasibilities of the CPA model,adaptive switching strategy,control parameters optimization algorithm,and the multi-mode control system are verified.Simulation results show the effectiveness of the adaptive switching strategy,while its chattering is weaker than the hard switching strategy.