Disturbance Rejection Control For Nonlinear Systems With Multiple Disturbances And Its Applications

Disturbances and uncertainties widely exist in industrial systems and inevitably bring adverse effects on performance and even stability of the closed-loop control systems.How to attenuate disturbances and uncertainties has therefore been an attractive subject in both control theory and practical engineering.Targeting at this old but still open issue,disturbance observer based control method provides an alternative practical solution and has also proved its effectiveness in plenty of industrial sites,e.g.,motion control systems and flight control systems.The conventional disturbance observer based control method mainly focuses on the estimation and attenuation of the lumped disturbances,ignoring analysis on the sources and features of the disturbances,which may limit its control performance.Toward this end,the dissertation focuses on the disturbanca rejection control for a of nonlinear systems with multiple disturbances,which are generated from die whole control looop,i.e.,the plant,actuator,and sensor.The main results and contributions are listed as follows:(?)For a class of motion control systems with multiple disturbancess,an optimized tracking controller is designed for the first time by combining nonlinear disturbance observer and continuous-time modelpredictive control method.The core is to predict the future tracking error and desired control input(including the lumped effects of disturbances/uncertainties)in the receding horizon by a higher-order sliding mode distturbance observer,which is designed based upon a rough nominal sy stem.Different from direct compensation approach in the most existing composite model predictive control methods,disturbance estimates are taken full advantage in the optimization.Experimental results of a permanent magnet synchronous motor servo system are presented to demonstrate the workability.(?)For a class of lower triangular systems with non-parametric uncertainties,a generalized dynamic predictive control scheme is presented for the first time by introducing a time-varying receding horizon.Instead of relying on the inherent robustness property of the standard predictive controller or on-/off-line parameter identi fication,a dual-Layer adaptive Law is designed to estimate the lumped effect of system uncertainties.Under a less ambitious but more practical control objecuve,namely the srmi-global stability,various nonlinearity growth constraints utilized in the existing related methods could be essentially relaxed,Nurrerical simulation and illustrative experimentot tests of a series elastic actuator system are provided to demonstrate both simplicity and effectiveness of the proposed method.(?)For a class of disturbed uncertain nonlinear systems with unknown output measurement errors,a new observer based controller is proposed for the first time Instead of inheriting from the estimation-error-driven structure of Luenberger type observer,the proposed disturbance observer only explicitly uses the control input.It has been proved that the proposed method endows the closed-loop system with strong robustness against output measurement errors and system uncertainties.With rigorous analysis under the semi-global stability critcrion,the guideline of gain dioice based upon the proposed structure is provided.Numerical simulation and comparative experiments of a helicopter model are implemented to better demonstrate feature and validity of the proposed method.(?)For permanent magnet synchronous motor servo systems,a new disturbance observer based control framework is designed for the first time by considering multiple disturbances,which is generated from the whole control loop,e.g.,cogging torques,load torques,friction torques,measurement error effects,dead-time effects,and pararneter perturbations.First of all,this section systematically analyzes several representative disturbances,particularly including their features and distribution in the practical servo system,and then,specifically puts forward a new disturbance rejection framework based on non-cascade structure.Comparative experimental results demonstrate that the proposed method achieves a better speed dynamic response and a higher accuracy tracking performance,even in the presence of multiple sources of disturbances.(?)For small fixed-wing unmanned aerial vehicles with additional flaps,a new dyaamic control allocation framework is proposed for the first time by integrating actuator dynamics.A disturbance observer based high-level tracking controller is first designed to accommodate the lag effect of the actuators and to oompensate the adverse effect of external disturbances.Then,a dynamic control allocator in low level is developed,which forces the actuator states to follow the optimal values of a receding-honzon performance Index.Compared to the conventional control allocation method that assumes ideal actuators with infinite bandwiddis,higher tracking accuracy of the UAV and better energy efficiency can be achieved by the proposed method.Simulations demonstrate the effectiveness of the proposed method in terms of output tracking and control allocation.