Research On Airship Station-keeping Control Method Based On Model Predictive Control

The multi-vectored propeller airship is a lighter-than-air aircraft with low flying speed.It is widely used in various military and civil missions for its vertical take-off and landing,long endurance station-keeping time and heavy loads,such as earth observation and communication relay.Using station-keeping mode,the airship can keep at the fixpoint for a long time with no energy consumption.Therefore,the station-keeping control system is the key to realizing station-keeping mode and it is very important.However,due to its particularity structure and working environment of the airship,the stationkeeping control system design is challenged by many factors.These factors including:The application of the conventional linear control theory to the airship is severely limited since its model is nonlinear and complicated.Some of its model parameters uncertainty is great since the flexible structure of the airship.The wind disturbance has a huge effect on its controller performance since its low flying speed.Its actuator(four multi-vectored thrusts)is severely limited and often operates in the saturated condition.The bad weather,heat flow,and wind disturbance easily lead to sensor damage,some states are difficult to measure.Due to the large volume and long distance between the controller at the payload and the thrusts at the euqational plane of the airship,the input time-delay is very large and is likely lead to degrading or even unstable of the control system.In this thesis,some front-line issues,such as external wind disturbance,actuator saturation,unknown states estimation and time-delay control problems of the airship are studied,and some model predictive control based composite controllers are proposed to solve the problems above and station-keeping is realized.The main works of this thesis are listed as follows:The nonlinear dynamics and kinematics model of the multi-vectored propeller airship are established,the mechanism of the aerodynamic and propeller force are analyzed in detail.Based on the nonlinear dynamics model,a T-S fuzzy model is established by linearization.An anti-wind disturbance and anti-saturation control system are proposed to solve the problem of strong wind disturbance and amplitude saturation of the thrusts.The control system is consists of three parts: model predictive controller,improved tracking differentiator and disturbance observer.The disturbance observer is mainly used to estimate the strong wind disturbance while the improved tracking differentiator is used to achieve the saturation suppression by planing the transition process.The proposed control system has an analytical control law,so the control law is obtained very fast with less online computation load.By including the disturbance observer and improved tracking differentiator,the control system could realize precision station-keeping under strong wind disturbance.Aiming at solving the actuator saturation(both amplitude and rate saturation)and estimating the unknown states,a station-keeping control system is proposed based on linear programming and enhanced Super-Twisting observer.The unknown states are estimated by the enhanced Super-Twisting observer,and by adding some new parameters in the traditional observer,the converge rate improved significantly.The linear programmingbased control allocation method can not only deal with the amplitude saturation but also rate saturation of the thrusts.A robust MPC-based station-keeping control system is proposed to deal with the input time-delay of the multi-vectored propeller airship.The station-keeping control is transformed into a path following control by coordinate transformation,and a guidance law is derived to obtain the desired velocity and angular velocity.A robust MPC feedback control law with time-delay is proposed to tracking the desired guidance law.A new Lyapunov function is put forward to obtain the upper bound of the time-delay min-max cost function of the robust MPC,and the corresponding linear matrix inequalities are derived to obtain the feedback gain.The proposed control system could stabilize the input time-delay system and drives the airship to the desired station-keeping position with high accuracy.This thesis studies on some station-keeping issues that the multi-vectored propeller airship might meet.Based on model predictive control theory,some integer control method are proposed or improved to dealing with disturbance rejection,anti-saturation,state estimation,and technical solutions time-delay control problem.The thesis provides a solid theoretical basis and technical solutions of station-keeping control of the airship.It also has significant value in engineering for station-keeping control system implementation.The proposed method is not limited to the multi-vectored propeller airships,it is also applicable to all the control object which have a similar disturbance,saturation,and time-delay.