Under the national strategy of “dual carbon” goals, the construction of a new type of power system with new energy as the main body has been put on the agenda. Wind power generation and photovoltaic power generation are important forms of new energy generation, so there is a lot of research on the integration of wind and solar power into the grid. However, wind and solar power sources are usually far from the load center and require long-distance transmission to connect to the grid. At the same time, as new energy continues to replace conventional power sources, the number of synchronous units supporting the basic operation of the grid continues to decrease, which can lead to a decrease in grid strength. The problem of stable operation of weak grids in new energy sources such as photovoltaic power generation is very prominent.
Photovoltaic power generation is generally connected to the grid through a voltage source type grid connected converter, so the control characteristics of the converter will dominate the grid connection characteristics of photovoltaic power generation. There is a significant control interaction between the weak current converter and the power grid, which easily leads to small disturbance stability problems characterized by broadband oscillation. A large amount of research has been conducted on this issue, mainly focusing on two methods: time-domain state space method and frequency-domain impedance method. The latter is widely used in new energy grid connected oscillations due to its ease of measurement, mature stability criteria, and black/gray box analysis for stability. Impedance modeling is the basis for impedance stability analysis. Reference [8] provides a method to establish an output impedance model considering the frequency coupling effect caused by the dynamic phase-locked loop under current loop control. Based on the generalized Nyquist criterion (GNC), the conditions for weak grid small disturbance instability are analyzed. On the basis of analyzing coordinate transformation and positive negative sequence decomposition, an improved sequential coordinate system is proposed, and the reasons for frequency coupling effect are analyzed using the improved sequential domain. The method of frequency scanning in the dq coordinate system and the improved sequential coordinate system is also provided. By using this method, it is possible to derive a difficult to analytically derive positive and negative sequence impedance model from the easily derived dq impedance through simple mathematical transformation and frequency translation, which facilitates impedance analysis in different coordinate systems (domains). We studied the impedance modeling method and validation in a two-phase stationary complex coordinate system, which is essentially another representation of the sequential coupling model. To better reveal the key mechanism of oscillation, reference [11] established a simplified model and analyzed the mechanism of solar inverter grid interaction resonance caused by phase-locked loop errors based on this. We established impedance models for grid connected converters in different dimensions and analyzed the impact of grid strength on stability. Through these methods, the conditions and parameter boundaries for stable operation of photovoltaic weak grids can be obtained, providing a theoretical and analytical basis for online stability control design based on parameter tuning and additional damping.
In grid connected stable adaptive control, the virtual impedance is adjusted by analyzing the remaining capacity of the solar inverter to achieve reasonable allocation of harmonic power in different solar inverters in the microgrid, and the adjustment range of the virtual impedance is determined based on stability. A self-adaptive droop control method is proposed to reduce the range of AC voltage waveform for the offshore DR-MMC hybrid DC power grid scenario by analyzing the steady-state transmission characteristics under constant voltage/frequency control and droop control. Improve the traditional feedforward link, adaptively adjust the gain coefficient of the feedforward link based on the impedance measurement results of the power grid, and enhance the robustness of the weak current network of the converter. Using the extended Kalman filtering method to identify the impedance of the power grid, and training an artificial neural network to judge system stability, optimizing the parameters of the current loop controller by reducing the search space of the current loop controller parameters, and improving the stability of the converter.
In addition to the analysis and control of small disturbances near the equilibrium point, the existence of equilibrium points in photovoltaic grid connected systems under non ideal power grids is also another key aspect of grid connected stability research. In this problem analysis, the dynamic control of the converter is usually simplified to better obtain the static operating range of the system. Studied the static stability problem caused by changes in the rate of change of the DC voltage loop, power loop, and current inner loop under different grid connection conditions. Furthermore, considering the influence of current constraints, the static power limit of new energy grid connected converters in weak grid conditions was discussed, and the static stability conditions of the system were given based on the rate of change of active power relative to active current in the DC voltage control loop of the converter. This analysis is very useful for setting active power reasonably in weak grid conditions to ensure the existence of static operating points. The above analysis indicates that under non ideal power grids, the active power of new energy grid connection will be limited, so it is necessary to cooperate with reasonable reactive power control. In this regard, the power transmission characteristics of photovoltaic power generation units under extremely weak power grids are analyzed, and a method of adaptive adjustment of droop control reference voltage is proposed to improve the active power transmission capacity. However, this method does not consider the stability constraints of weak power grids, and the effective release of active power capacity cannot be guaranteed.
In summary, it can be seen that the existence of static equilibrium points (such as active power boundaries under multiple static constraints) and small disturbance stability control near the equilibrium point are two important aspects that affect the active power output capacity of photovoltaic grid connection. However, further research is needed to develop control methods that meet both requirements. Therefore, this article proposes a voltage reactive droop coefficient (Q-V droop coefficient) adaptive adjustment method, which can ensure the stable output of given active power of solar inverters under different weak grids. Firstly, introduce the process of optimizing the Q-V sag coefficient in one step, analyze the method of solving the Q-V sag coefficient based on the working point range under the limitation of grid voltage and current, provide the optimization results of the Q-V sag coefficient and the results of the feasible range changing with the strength and output power of the power grid, and introduce its specific algorithm implementation. Then, the Q-V sag coefficient secondary adjustment is introduced, including small disturbance analysis, impedance modeling, stability analysis of photovoltaic grid connected systems, and the method of using ANN to fit the pole lines of interconnected systems to determine system stability. The implementation of the Q-V sag coefficient secondary adjustment algorithm is also introduced. Secondly, introduce the impedance identification method used in this article to achieve the applicability of the power grid. Finally, the effectiveness of the method was verified by algorithm implementation and simulation analysis on a wide real-time simulation platform.