Research On Improved High-order Slide Mode Control And Allocation System For Reusable Launch Vehicles

Reusable launch vehicles(RLV)are widely used in both civil and military purpose,and have become a popular research topic for researchers at home and abroad.RLV has a wide range of flight and the aerodynamic environment changes drastically.These characteristic features demand a lot for the adaptability and robustness of the control system.The ability to adjust control performance autonomously is the current trend in the field of flight control.This paper mainly studies the control framework of the unpowered aircraft in the reentry phrase,and the main content of research are showed as follows.First of all,the common-used coordinate system and related coordinate transformation methods are introduced.Based on the dynamic and kinetics theory,the six-degree-of-freedom mathematical model of aircraft id derived.Under certain premise assumptions,the attitude control model and control allocation model of the RLV are constructed for the subsequent control design.The basic framework laid a solid foundation for further research.Secondly,for the attitude control design of RLV,according the the multi-time scale theory,the control system can be divided into two subsystems which are inter-loop and outer-loop respectively,and adaptive integral terminal sliding mode controller are used in both.Considering the external disturbance of torque and the uncertainty of moment of inertia,this paper designed a super-twisting disturbance observer to estimate the uncertainty.For the actuator saturation,advanced dynamic anti-windup compensation scheme is developed,which effectively shortens the control saturation time.The simulation results verified the robustness and effectiveness of the designed control scheme.And then,for the virtual torque command generated by the sliding mode controller,by introducing the concept of control allocation,the combined framework of controller design and multi-actuator control allocation is established.The control allocation strategy based on the improved firework algorithm is designed.By using chaotic initialization and chaotic spark method,the space of solution can be effectively searched.According to the phenomenon of fireworks clustering,the K-nearest neighbor algorithm is used to estimate the clustering density,and the blast radius will adaptively adjusted to reduce the clustering effect.Then the chaotic firework algorithm is improved by introducing the idea of various groups.A multi-group collaborative search strategy based on the beetle antennae search algorithm is proposed to solve the control allocation problem.The simulation results show the proposed algorithm is better than the tradition firework algorithm and particle swarm optimization algorithm.Next,considering the rate constraint of actuator,the uncertainty of control efficiency matrix and the requirement fo real-time control,a robust fast control allocation strategy based on improved convex optimization is proposed.The traditional inverse method and quadratic programming are introduced,and then a mathematical optimization problem based on robust optimization is established for the performance index with the smallest deflection of actuator and Huber loss function is used to replace the traditional two-norm.Last the robust optimization problem can be converted into a second-order cone scheme.The simulation results showed that the proposed method can quickly complete the task of allocate the control torque command in the presence of uncertainty,and the tracking performance is good`.Finally,this paper draws the final conclusion,and summarized the main content of research.Also the shortcomings and further research direction are analyzed and discussed.