With the advancement of communication technologies and the development of the smart grid, today's physical power systems present an ever-growing dependency on cyber resources. It increases cyber vulnerabilities, causing safety and system stability concerns. In this article, a data-driven resilient automatic generation control (AGC) scheme is proposed under a false data injection attack (FDIA). The key idea is to identify the relationship between AGC signals and system operational conditions. This is achieved by the proposed regression-based FDIA signal predictions, including sequence-to-point prediction and the long short-term memory network-based prediction. It allows us to reconstruct the AGC control signals without being affected by FDIAs and to attenuate attacks in the closed control loop, thus alleviating the impact of FDIA on system performance. Numerical results carried out on the benchmark systems validate the effectiveness of the proposed method.

In contrast to the traditional high-voltage direct current (HVDC) based offshore wind farms, series-connected offshore wind farms do not require transformers, voltage source converters (VSCs), and offshore platforms, thereby significantly decreasing costs. In this study, the onshore converter, which is based on the half bridge (HB) modular multilevel converter (MMC) and full bridge (FB) MMC, is proposed to reduce the costs and power loss of the series-connected offshore windfarm. In this type of onshore converter, the DC voltage of the HB-MMC is controlled to be the rated DC voltage, while the DC voltage of FB-MMC can vary from negative rated DC voltage to positive rated DC voltage. With the variable DC voltage of the FB-MMC, the DC link current of the series-connected offshore windfarm can be controlled at will. Moreover, the DC link current control strategies is proposed to reduce the power loss of the offshore windfarm. The capacitor voltage balance control, which is suitable for FB-MMC with variable DC voltage, is proposed to ensure the normal operation of the onshore converter. Finally, simulation and experimental results verify the feasibility of the proposed onshore converter and the corresponding control algorithm.

This paper compares three methods, the impedance model based analysis, the positive net damping analysis and the open-loop modal resonance analysis. The differences and connections among three methods are revealed, which helps to gain better understanding on the limitations of the methods for studying the mechanism of sub-synchronous oscillations (SSOs) caused by grid-connected wind turbine generators (WTGs) in a power system. Following useful conclusions are obtained, which are evaluated and demonstrated by study cases: (1) Both the negative resistance and negative net damping are highly related with strong dynamic interactions between the WTGs and AC grid under the condition of modal resonance. This explains why resistance and net damping may become negative from the standpoint of system modal condition. (2) It is possible that results of IMA and PDA may not reflect correctly the system SSO instability risk brought about by the WTGs. The reason is that the IMA and PDA are based on the sufficient condition of system stability. Assessment made by the IMA or PDA may be very conservative. In study cases, three methods are compared in the parameter space of converter controllers of the WTGs such that all the parameter conditions are covered in comparison.

This paper analytically derives the small-disturbance stability limit of a grid-connected wind farm with multiple PMSGs in the timescale of DC voltage dynamics. The derived stability limit reveals the mechanism about how the dynamics of DC voltage control and the PLLs may jointly affect the small-disturbance stability of the wind farm. Flagging the impact of PMSGs by a computational function, the stability limit indicates that decreasing value of flagging function or/and increasing maximum eigenvalue of network reactance matrix implies increased instability risk. Particularly, increasing the loading conditions and number of PMSGs in the wind farm or/and the length of transmission line connecting the wind farm with main grid may make the wind farm more prone to the loss of small-disturbance stability. Although the stability limit is derived under the assumption of identical dynamics of PMSGs, the derived small-disturbance stability limit can be applied to approximately assess the small-disturbance stability of the wind farm in the timescale of DC voltage dynamics, where dynamics of PMSGs may be either similar or considerably different. The approximate small-disturbance stability assessment is simple because the assessment does not need to establish the parametric model of the wind farm.

In this study, the DC voltage oscillation mechanism is investigated based on the linearized models of a voltage-source converter (VSC)-based multiple terminal DC (MTDC) system. First, the mechanism of DC voltage oscillation generation is revealed. DC voltage or DC voltage droop-controlled VSCs are identified as the DC voltage oscillation sources when master–slave or DC voltage droop control is separately applied to the VSC-MTDC system. Meanwhile, the active power-controlled VSC is equally regarded as a constant power source that does not provide any oscillation to the VSC-MTDC system. Subsequently, the impact of the dynamics of VSCs and MTDC network on the DC voltage oscillation is separately investigated. It is theoretically verified that the negative eigenvalues of the MTDC network conductance matrix can reduce the stability of the DC voltage oscillation. The constant power load is identified as the physical mechanism of a problematic scenario that induces DC voltage oscillation in the generalized VSC-MTDC system. Moreover, it is demonstrated for the first time that the steady-state power flow may cause DC voltage oscillations when the power flow is reversed. Finally, two VSC-MTDC systems, which individually adopt master–slave and DC voltage droop control, are presented to validate the findings using the standard MATLAB SIMULINK model.