Research On Multi-source Disturbance Estimator-based Robust Flight Control And Its Applications

Multi-source disturbances consisting of input disturbances,model uncertainties,and measurement errors are inevitably encountered by aerial vehicles under complex flight conditions.Systematic consideration of the active disturbance rejection in the control de-sign is essential to guarantee the flying qualities and flight safety of aerial vehicles.How-ever,the classical control theory indicates that the simultaneous rejection of multi-source disturbances often comes with refined trade-offs,rendering the design process sophisti-cated and challenging.The robust trajectory-tracking problem for control systems in the presence of multi-source disturbances is addressed in this dissertation.On the basis of the nominal tracking control,a unified disturbance estimator-based robust control framework is proposed by structurally combining different types of disturbance compensation mechanisms.This framework is capable of rejecting disturbances in measurement and control systems,and it is further applied to and verified by several aerial vehicle platforms.The novelties of this dissertation are summarized as follows.The robust trajectory-tracking problem of velocity-sensorless systems is addressed.Two uncertainty and disturbance estimator(UDE)-based input disturbance compensation approaches are proposed.In one approach,the feasible relative order of the filter in UDE is derived such that the physical realizability of UDE is ensured,and the disturbance can be estimated and compensated by UDE under the velocity-sensorless condition.The velocity-tracking control is delivered by the passivity technique,which injects damping to the system to replace the unavailable velocity feedback control.A two-way coupled structure of the Luenberger state observer(LSO)and UDE is proposed in the other ap-proach,where LSO provides UDE with the velocity estimate,while UDE compensates for the effects of disturbance on LSO.The two approaches are applied to the 3-DOF he-licopter.Both simulation and experimental results show that the proposed approaches can achieve not only disturbance estimation and compensation but also high-accuracy reference-tracking.A novel measurement and control system model-based measurement error estima-tor(MEE)is proposed to resolve the robust trajectory-tracking problem for systems with sensor performance constraints.Compared with classical filtering-based methodologies,MEE achieves accurate online estimation and active compensation of the measurement error in the control system,by maximizing the utilization of not only the models of the sensors,controller,and control plant but also the real-time system input and output infor-mation.Then,a pre-filter is introduced to pre-process the measurement signal,such that the bandwidth requirement on MEE is relaxed,and the sensitivity of MEE to model uncer-tainties is thus reduced.Furthermore,a novel sensor lag compensator(SLC)is designed using the measurement and control system model.SLC is embedded in the controller to dynamically reconstruct and compensate for the sensor-inertia-induced dynamic lag con-sidered in the complex sensor model.The applications of the proposed control approaches to the 2-DOF helicopter are presented.Comparative simulation and experimental results verify the performance advantage of the proposed approaches over the Kalman filter-based approach.The robust trajectory-tracking control of a class of second-order systems under multi-source disturbances is studied.A multi-source disturbance estimation framework is pro-posed.The novelty of this framework lies in the utilization of partial accurate model infor-mation and partial accurate state measurements to construct MEE and UDE,respectively,which help the closed-loop controller simultaneously compensate for multi-sources distur-bances.Besides,a singular perturbation parameter ? is introduced in the proposed frame-work to tune the bandwidths of the two estimators,and singular perturbation theory-based analysis shows that reducing ? enhances the multi-disturbance rejection performance and improves the closed-loop system stability.The proposed MEE+UDE-based disturbance compensation approach is applied to the 2-DOF helicopter and quadrotor,and it exhibits superior performance over classic robust control approaches in the comparative simula-tions and experiments with respect to the transient control accuracy,steady-state control accuracy,and ease of parameter tuning.The extension and application of MEE-based control to the robust flight-path angle synchronized tracking of fixed-wing aircraft in the presence of measurement errors are in-vestigated.By considering the non-minimum phase characteristics existing in the fixed-wing aircraft flight-path angle dynamic model,a three-module-based robust distributed control approach is proposed by integrating the MEE-based measurement error compen-sation methodology with the classical distributed observation technique and non-minimum phase system control technique.This approach is featured with the decoupled principle in both structural and functional aspects,which allows the improvement or redesign of any functional module to meet practical requirements without affecting other modules.Sim-ulation results for a formation of F-16 aircraft show that the proposed approach delivers robust synchronized tracking of a time-varying flight-path angle reference.