In the 21st century, global warming has emerged as one of the most pressing challenges faced by humanity. This phenomenon not only poses a threat to the earth’s ecosystems and biodiversity but also exerts profound impacts on the global economy, social stability, and human health (Gernaat et al., 2021; Xie et al., 2024a; Sun et al., 2021a; Xie et al., 2024b; Sun et al., 2023). In response to this crisis, the international community is actively engaged in the pursuit of sustainable solutions, among which enhancing the proportion of renewable energy generation as a substitute for conventional fossil fuel-based power generation has emerged as a key strategy (Olabi and Ali Abdelkareem, 2022; Mai et al., 2014; Haegel and Kurtz, 2022). According to Renewable Capacity Statistics 2024, the global additional installed capacity for photovoltaic (PV) power generation in 2023 reached 345.5 GW, while the additional installed capacity for wind power generation was 115.97 GW. In comparison to the year 2022, the year-on-year growth rate of the installed capacity for photovoltaic power generation was 22%, and for wind power generation, it was 9% (International Renewable Energy Agency (IRENA), 2024).

With the rapid proliferation of renewable energy generation within power systems, the inherent volatility and randomness of these new energy sources pose significant challenges to the secure and stable operation of the grid. voltage source converters (VSCs), as pivotal interfacing devices between renewable energy sources and the power grid, have garnered considerable attention (Li et al., 2016; Sun et al., 2021b; Xiao et al., 2024a; Martínez-Parrales et al., 2021). To address the aforementioned challenges, numerous VSC control strategies have been proposed, which can be broadly classified into two categories: GFL and GFM controls (Xiao et al., 2024b; Bhatta et al., 2024; Ducoin et al., 2023). GFL, lacking the capability to establish a grid, finds it difficult to provide support to the power network. In contrast, GFM, independent of phase-locked loop (PLL) for generating phase angle, possesses the ability to form an independent power grid, thus attracting increasing scrutiny. Among the various GFM control strategies, the most representative are the virtual synchronous generator (VSG) control and the droop control method (Verbe et al., 2021; Li et al., 2020; Song et al., 2021; Li et al., 2017).

While VSG emulates the characteristics of a synchronous generator through control means, enabling functions such as primary frequency modulation, damping, inertia support, and voltage regulation to provide frequency or voltage support to the system, it is also capable of directly responding to the rate of change of frequency (ROCOF) in the system, which is a primary advantage of VSG control over droop control. Despite some scholars summarizing the differences between the two (Chen et al., 2023; Mohammed et al., 2024; Meng et al., 2019), a detailed comparative analysis of their control characteristics and support capabilities for the system is still lacking.

For the existing research problems, this paper conducts a comparative analysis of the control characteristics of VSG and droop control within a nine-bus system based on 100% renewable energy generation. The contributions of this paper could be summarized as.

(1) The typical control architecture and control principle of GFL and GFM are introduced, and the existing typical control methods of GFM and GFL are summarized.

The remainder of the paper is organized as follows: Section 2 presents the principles of VSG and droop control. Section 3 conducts the small-signal modeling. Section 4 designs the simulation experiments and analyzes the simulation results. Section 5 provides the conclusions.

The synchronization mode of the GFL is to measure the phase information of the parallel node through the phase-locked loop as the reference of the reference voltage phase, so as to realize the “following” of the power grid, which generally shows the characteristics of current source. The GFM must rely on a strong AC power grid to work normally. When the power grid strength is weak, the output current has a great impact on the terminal voltage, which is easy to lead to the instability of the grid following power supply under large/small disturbances. The synchronization mode of grid converter is similar to that of synchronous machine. It adjusts the frequency of reference voltage through power exchange with the grid, so as to achieve synchronization with the grid. On the whole, it shows voltage source characteristics externally. Network type control does not need phase locked loop, and has the ability to independently construct voltage and frequency. It can be applied to provide stable AC voltage in weak current network or even isolated network scenarios.

The typical GFL control for power calculation is shown in Figure 1, which can be mainly divided into three parts: PLL, power outer loop current inner loop double closed loop control, and power calculation. PLL tracks and combines the voltage phase of the node, generates the phase angle, and converts three-phase AC electrical quantities such as voltage and current into two-phase direct flow through Parker transform to calculate the active and reactive power output by VSC. The accurate control of VSC output active and reactive power is realized through the power outer loop, and the current inner loop ensures that VSC will not have overcurrent. Control performance can be improved by adding additional controls such as droop control.

The biggest difference between GFM and GFL is that GFM does not rely on PLL to realize grid connection. The control architecture can be divided into three parts: phase angle generation, power calculation, voltage outer loop current inner loop double closed loop control. The actual output power is obtained through power calculation, and the phase angle is generated by the difference between the actual output power and the target power and the set active power and frequency coupling relationship. The target voltage is generated through the double closed-loop control of voltage outer loop and current inner loop. It is worth mentioning that on this basis, additional control such as reactive power voltage droop can be added to improve the control performance. The typical control architecture of GFM is shown in Figure 2.

At present, typical VSC control can be divided into GFM and GFL. GFM control includes typical VF control, VOC Control and GFM control based on PSL, among which VSG and droop control are the most widely used. GFL control includes PQ control, droop control, VSG control and other GFL controls with additional control strategies. As shown in Figure 3.

The GFM control structure employing droop/VSG control is depicted in Figure 4. On one hand, the filter capacitor voltage and the filter inductor current are subjected to the Park transformation, converting the three-phase AC quantities into two-phase DC quantities. On the other hand, these quantities are fed into the power calculation module to compute the active and reactive power outputs. The active power (P), reactive power (Q), active power setpoint (P _set ), and reactive power setpoint (Q _set ) are then delivered to the GFM outer loop control (VSG/droop control) to generate the phase angle (θ) and voltage amplitude (u _td *). After passing through the classical double-loop control comprising an outer voltage loop and an inner current loop, the target voltages (u _ref,dq ) are produced. These are then transformed back to the three-phase target voltages (u _ref,abc ) via an inverse Park transformation, and finally, pulse-width modulation (PWM) is utilized to generate the gate control signals, thereby completing the control process.

The development of small-signal models for both droop control and VSG control is essential for elucidating their control characteristics and discerning their principal differences. As illustrated in Figures 7, 8. These models facilitate a detailed analysis of the dynamic responses under small perturbations, which is pivotal for the design and optimization of control algorithms within a power system. By comparing the small-signal models of droop control and VSG, one can gain insights into the inherent advantages and limitations of each control strategy. This analysis is particularly relevant in the context of the evolving electric grid, where the interplay between traditional and renewable energy sources necessitates sophisticated control mechanisms to ensure seamless and stable power delivery. The VSC grid connection topology is shown in Figure 9, where E is the VSC voltage and U _g is the voltage of the parallel node.

A simulation based on a nine-bus system powered entirely by renewable energy sources was conducted, with two operational scenarios designed: a sudden load increase and a power source disconnection. The system topology is illustrated in Figure 10. VSC1 was configured to employ either droop control or VSG control. The control characteristics and differences between the two were analyzed.

In the initial state, VSC1 employs a GFL control strategy with droop control, while VSC2 utilizes a GFM control strategy with droop control. VSC3 has the option to operate under either a GFM control strategy with droop control or a VSG based GFM control strategy.

In conclusion, this paper has delivered a comprehensive comparison of VSG and droop control within a fully renewable energy-powered nine-bus system. The study has highlighted that VSG control offers enhanced damping, greater stability in frequency and voltage, and crucial inertia support, outperforming droop control. These findings are significant for power system stability, especially with the increasing integration of renewable energy sources. The insights from this research can guide future developments in power system control, emphasizing the potential of VSG control in maintaining grid reliability amidst the energy transition.