Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.755476
Title: Advanced control of doubly-fed induction generator based wind turbines for dynamic performance improvement
Author: Shi, Kai
ISNI:       0000 0004 7428 4704
Awarding Body: University of Liverpool
Current Institution: University of Liverpool
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The Doubly-fed induction generator (DFIG) is one of the most popular generators used in variable speed wind turbines with the merits of high efficiency and using small-scale power electronic converter. To improve the reliability and stability of power grids with high penetration of wind power, low-voltage ride through (LVRT) requirements for wind turbines have been introduced by many countries in their grid codes. Wind turbines are required to stay connected to the grid during voltage dips and provide reactive power support. The standard vector control (VC) is challenged to fulfill those LVRT requirements as it does not provide a satisfactory transient response for DFIG-WTs. The DFIG is sensitive to grid disturbances as its stator terminals are directly connected to the power grid. Using advanced control strategies is a cost-effective way that can improve the dynamic performance of the DFIG-WTs based on existing hardware. This thesis aims to develop control algorithms for DFIG-WT in order to improve the dynamic performance of DFIG-WT and enhance the its robustness against uncertainties and disturbances. The performance of VC with different controlled variables of DFIG has been evaluated at first. An external-loop for stator flux is proposed to accelerate the damping of stator flux and improve the terminal voltage control during voltage dip. Performance of the proposed controller is verified via simulation and experiment tests in comparison with the standard VC-based active/reactive power control and terminal voltage control. As VC can only provide asymptotically decoupled control of active and reactive power due to its constant assumption of stator voltage and flux, feedback linearisation control (FLC) is designed for DFIG to provide a fully decoupled control and maintain a consistent control performance among different operation points. The FLC of different controlled variables have been proposed, and the transient response of DFIG with those FLCs during voltage dips has been investigated by simulation. The FLC requires accurate system model and parameters, and full-state feedbacks are needed. The parameter uncertainty could strongly degrade the performance of FLC. The disturbance observer based FLC (DOFLC) is proposed to improve its robustness against parameter uncertainties, but only constant or slow timevarying disturbances can be handled. A perturbation estimation based nonlinear adaptive control (NAC) for the decoupled control of the active and reactive power of DFIG is proposed. Perturbation observer (PO) is designed to estimate fast timevarying perturbations, including system nonlinearities, parameter uncertainties, unmodelled dynamics, and unknown disturbances. With compensation of system perturbations, the fully decoupled control of active and reactive power is realised. The dynamic performance and robustness against uncertainties and disturbances is verified via both simulation and experiment tests in comparison with the VC and DOFLC. Perturbation estimation based NAC is applied to the speed and terminal voltage control for the DFIG-WT. Choosing the rotor speed and stator flux as system outputs, two second-order-represented input-output linearised relationship with relative degree 4 are obtained. Two third-order perturbation observer are designed for the second-order subsystems. Time-varying perturbations, including all system nonlinearities, unmodelled dynamics, and unknown disturbances, can be estimated as a lumped term. Simulation results are carried out to verify the performance of the proposed controller during voltage dips and voltage swells and the robustness against parameter uncertainties. Unbalanced grid voltages have frequently occurred due to an imbalance of line impedances, unbalanced loads, and some asymmetric grid faults. The conventional solution with VC under unbalanced grid conditions requires four control loops in both positive- and negative- sequence dq reference frame. A Perturbation estimation based NAC for the power control to DFIG operation under unbalanced grid conditions is developed in stationary aß references frame, with which the separated control in positive and negative sequence dq reference frame is not required. Perturbation observer is designed to estimate sinusoidal time-varying perturbations. By compensating those perturbations with PO, tracking control of the rotor currents with linear output feedback controller is realised. To eliminate the active power oscillations when grid voltage is unbalanced, rotor current references are given based on the current-power relationship modelled under unbalanced grid conditions. The results presented in this thesis address the problems mentioned above. When the stator flux is controlled during a voltage dip, the DFIG shows improved performance in damping the oscillated natural flux and maintaining a good terminal voltage profile. FLC based DFIG controllers are designed to achieve the fully decoupled control of the DFIG such that better dynamic performance is obtained than using the VC. To overcome the drawbacks of FLC on its weak robustness of the FLC against uncertainty, perturbation observer based nonlinear adaptive control is proposed so the dynamic performance of the DFIG can be improved when there are parameter uncertainty, unknown dynamics, and external disturbances, etc.
Supervisor: Jiang, Lin Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.755476  DOI:
Share: