Inverter control for wind–solar–storage based on distributed generation and virtual synchronous machines
- 1Ningxia Branch, PowerChina Renewable Energy Co., Ltd., Yinchuan, China
Abstract
To enhance the operational stability of the power system and address the issues of weak inertia support and sluggish dynamic response of traditional inverter control strategies when dealing with complex scenarios involving a high proportion of new energy sources, a new intelligent control strategy has been proposed. This study adopts the VSG control strategy as the core technology for the regulation of wind-solar-storage power generation inverters, and divides the dynamic response process into acceleration period, deceleration period, reverse acceleration period and reverse deceleration period. Different control strategies are adopted for different system states at different stages. Additionally, the DDPG algorithm is introduced, and the single DDPG agent is extended to a multi-agent interactive deep deterministic policy gradient algorithm. Finally, the multi-agent collaborative optimization is achieved in a “centralized training and decentralized execution” form. The research results show that in the case of sudden load increase, the voltage deviation at the grid connection point is stabilized at 371.7 V, and the voltage recovery time is significantly shorter than the comparison strategy. At the same time, the time-weighted absolute error integral reaches optimal values of 1.82 V/s2 and 1.85 V/s2 in scenarios of sudden load increase and sudden drops in photovoltaic power, and the strategy efficiency reaches 97.83%, and the steady-state accuracy reaches 99.63%, demonstrating superior control capabilities. In the three typical fault conditions of practical application, the average frequency deviation of the control strategy was 0.197 Hz, the number of protection lockouts was 0.3 times, the average depth of voltage sag was 24.63%, the average recovery time was 1.21 hours, and the number of abnormal warnings did not exceed 5 times. The proposed framework for coordinated wind, solar, and storage regulation, based on distributed power generation and virtual synchronous machine, can effectively support the safe, flexible, and efficient operation of the new power system under high penetration of renewable energy.
Related
- On the Application of Hybrid Mixed Domain Decomposition Methods to Permanent Magnet Synchronous Machines
- Representative Learning for Distributed Support Vector Machines
- Concurrent Parameter Learning and Current Control for Large-scale Grid-following Inverter-based Resources
- Safety Verification for Voltage Control in Inverter Dominated Radial Distribution Networks under Actuator and Sensor Adversarial Attacks
- Realisation-Aware Fractional Shaping for Robust Current Control of a Frequency-Controlled Parallel-Resonant Inverter
BibTeX
@article{cao2026inverter,
title = {{Inverter control for wind–solar–storage based on distributed generation and virtual synchronous machines}},
author = {Heng Cao},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0359162},
url = {https://doi.org/10.1371/journal.pone.0359162},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Inverter control for wind–solar–storage based on distributed generation and virtual synchronous machines
AU - Heng Cao
PY - 2026
DA - 2026/10/27
JO - PLOS One
DO - 10.1371/journal.pone.0359162
UR - https://doi.org/10.1371/journal.pone.0359162
AB - To enhance the operational stability of the power system and address the issues of weak inertia support and sluggish dynamic response of traditional inverter control strategies when dealing with complex scenarios involving a high proportion of new energy sources, a new intelligent control strategy has been proposed. This study adopts the VSG control strategy as the core technology for the regulation of wind-solar-storage power generation inverters, and divides the dynamic response process into acceleration period, deceleration period, reverse acceleration period and reverse deceleration period. Different control strategies are adopted for different system states at different stages. Additionally, the DDPG algorithm is introduced, and the single DDPG agent is extended to a multi-agent interactive deep deterministic policy gradient algorithm. Finally, the multi-agent collaborative optimization is achieved in a “centralized training and decentralized execution” form. The research results show that in the case of sudden load increase, the voltage deviation at the grid connection point is stabilized at 371.7 V, and the voltage recovery time is significantly shorter than the comparison strategy. At the same time, the time-weighted absolute error integral reaches optimal values of 1.82 V/s2 and 1.85 V/s2 in scenarios of sudden load increase and sudden drops in photovoltaic power, and the strategy efficiency reaches 97.83%, and the steady-state accuracy reaches 99.63%, demonstrating superior control capabilities. In the three typical fault conditions of practical application, the average frequency deviation of the control strategy was 0.197 Hz, the number of protection lockouts was 0.3 times, the average depth of voltage sag was 24.63%, the average recovery time was 1.21 hours, and the number of abnormal warnings did not exceed 5 times. The proposed framework for coordinated wind, solar, and storage regulation, based on distributed power generation and virtual synchronous machine, can effectively support the safe, flexible, and efficient operation of the new power system under high penetration of renewable energy.
SN - 1932-6203
ER -