| As a typical high order and multi-input multi-output (MIMO) system, the interaction between converters of voltage source converter multi-terminal DC (VSC-MTDC) is very strong, even though each converter is individually well-designed. And the disturbance of each converter, such as frequent fluctuations of new energy source, periodic change of passive load, may cause the shock of power flow and DC voltage, even the other converters and ac systems could be affected. Furthermore, the interaction may degrade the stability and dynamic performance of whole system and cause instability. As a result, the converters which depend on stability voltage of common connection point, such as passive network, new energy, and weak system, are adversely affected. This paper aims to identify the coupling mechanism, reveal reasons and influence factors. Then, decoupling control of VSC-MTDC has been studied to weaken the interaction between converters and make the complex high-order system into combination of multiple low-order systems. Finally, the robustness and decoupling performance of converter are optimized.Firstly, this paper aims to revise and improve the defect of existing small signal model of VSC-MTDC through theoretical analysis and derivation. Eliminating the impact of ac system short-circuit capacity to accuracy of the small signal model by choosing d-q synchronous reference frame flexibly. Then according to PWM modulation principle, the coupling mechanism between ac and dc of converter has been analysis and used to optimize the linear equation of converter. Meanwhile, the linear equation of controller, and DC network have been deduced. Finally combining these linear equations, the global small signal model of VSC-MTDC has been proposed. Then By comparing of Electromagnetic transient simulation and the small-signal model simulation, the small-signal model has high accuracy and reference value for stability analysis, coupling mechanism and control design.Stability analysis is the premise of coupling mechanism research and control design. Modal analysis method and correlation factor matrix have been used to analysis the stability and interactions between state variable of VSC-MTDC. Then, in order to identify the crux of instability and Stability optimization, a sensitivity analysis method has been proposed to reveal the relationship between characteristic values and System parameters and control parameters.Based on small signal model, stability analysis, and small signal analysis method, the coupling mechanism of VSC-MTDC have been studied in two aspects. The one is the interaction between converters without dc voltage control. Another is the interaction between dc voltage control converter and others. The study has shown that the interaction is produced by DC voltage coupling between converters and DC-AC coupling between AC side and DC side of converter. And the influence factors contains DC voltage, DC line impedance, converter DC capacitance, converter modulation, and the active power absorbed by converter.Based on coupling mechanism analysis and state variable feedback decoupling theory, the decupling control with or without communication methods of VSC-MTDC have been studied. Three kinds of decoupling control was proposed, such as control based on the local DC voltage feedback, the control based on the local DC current feedback, and control based on the wide-area DC voltage feedback. Proved by simulation, all of these controls can weaken the interaction and realize decoupling between converters.Finally, the robust control strategy for VSC-MTDC has been proposed. In the case of consideration of converter uncertainty, the robust stability, robust performance, and decoupling performance of converter have been set as the optimization target. Based on multivariable feedback control theory such as H∞ mixed sensitivity design and Internal model control, the active power, reactive power control and ac voltage control of converter have been design. The simulation shows that these control have better robustness and decoupling performance than diagonal proportional-integral control. |
PDF Full Text Request