Robust Control Approaches For Uncertain Systems And Their Applications In Power Systems

In China,coal-fired power plants and hydro power plants play the dominant role for electricity supply,supplemented with the renewable distributed energy.As known,the renewable distributed generators represent the developing trend of future clean energy.To alleviate the increasing energy and environmental crisis,key technologies of both clean high-efficiency coal-fired units and distributed generators based on renewable energy are urgent to be developed.Recently,advanced control methods are widely studied and applied to enhance the operation of coal-fired units and distributed generators in a microgrid.Therefore,in this paper,some key issues on robust control methods are investigated and applied in power systems to obtain satisfactory control performance.The main contributions of this paper are summarized as follows:(1)Two robust control approaches based on the uncertainty and disturbance estimator(UDE)are investigated for systems with mismatched uncertainties.Firstly,stabilizability conditions are studied for the UDE-based control system,which renders a systematic design method for the reference model.Then,an asymptotic tracking controller is proposed to handle both the matched and mismatched uncertainties.With the integration of UDE and sliding mode control(SMC),a UDE-based chattering-free sliding mode control approach is proposed for a class of nonlinear systems with mismatched uncertainties,which eliminates the chattering phenomena in conventional SMC.(2)A novel sliding mode control approach based on the adaptive feedback linearization is studied for affine-nonlinear systems.To eliminate the linearization errors caused by modeling mismatch,the adaptive feedback linearization method is proposed.To reduce the chattering phenomena,a super-twisting second-order sliding mode controller is designed based on the linearized models of nonlinear plants.Moreover,closed-loop stability is strictly proved.To illustrate the effectiveness of the proposed approach,it is applied to the coordinated controller design for boiler-turbine systems with uncertainties.To obtain key features of subcritical boiler-turbine systems,nonlinear control-oriented models are developed based on mechanism analysis and model reduction techniques.The established models are validated by the history operation data of a certain 300 MW real coal-fired units.(3)With the integration of predictive control and SMC,a dual-mode sliding mode predictive control approach is investigated for nonlinear systems with uncertainties and input constraints.While the system states are outside of the sliding mode region,the optimal control sequence generated by the receding-horizon predictive controller is adopted,where the objective function contains tracking errors,control input and the sliding mode errors.While the system states are inside of the sliding mode region,the discrete sliding mode control law is applied to obtain fast response and disturbance rejection.Furthermore,the input-to-state stability of the closed-loop system has been proved.Considering the limited operation conditions of a practical coal-fired unit,a sliding mode predictive controller is designed for the ultra-supercritical unit to achieve coordinated operation of the boiler and turbine.(4)The robust control approaches are studied for the power converters in distributed generations.Firstly,the UDE-based chattering-free sliding mode control method is applied for the voltage control of DC-DC power converters to deal with the source voltage variations,load change and parameters uncertainties.Then,the virtual impedance control method is proposed for the parallel inverters to enhance the stability robustness,power quality and dynamic performance of power regulation.The techniques of coprime factorization and all stabilizing controller is introduced for the virtual impedance design,which theoretically guarantees the closed-loop stability of parallel-operated inverters in the microgrid.(5)To deal with the synchronization of distributed generators in a microgrid,two estimation approaches are investigated to quickly obtain the accurate parameters of the grid voltage.An improved sinusoid-locked loop is proposed by introducing a virtual stator impedance,to eliminate the undesired equilibrium point.Therefore,good estimation performance could be guaranteed in the presence of large voltage variations.To obtain fast and robust estimation in the presence of harmonic disturbances,a sliding mode observer is designed based on the exosystem model.For the islanded networked microgrids,the coordinated control strategy of multi-generators and multi-microgrids are investigated.A unified two-layer framework of multi-agent-based distributed secondary control of microgrids is proposed.The lower layer is responsible for the coordination of multiple generators in an individual microgrid,to handle the task of frequency/ voltage restoration,power sharing and economic operation.The upper layer is responsible for the coordination of multiple microgrids in a large region,to achieve the intentional islanding,re-connection and power sharing of microgirds.The proposed two-layer control strategy provides a reliable solution for the flexible operation of microgrids between different modes.