Embedded Machine Learning for Solar PV Power Regulation in a Remote Microgrid
This addresses power regulation challenges for remote microgrid operators by reducing delays, though it is incremental as it applies existing ensemble learning methods to a new deployment context.
The paper tackled solar inverter power regulation in remote microgrids by deploying trained machine learning models on embedded edge devices to reduce communication delays, achieving inference times of about 0.1ms per input with results matching desktop PC performance.
This paper presents a machine-learning study for solar inverter power regulation in a remote microgrid. Machine learning models for active and reactive power control are respectively trained using an ensemble learning method. Then, unlike conventional schemes that make inferences on a central server in the far-end control center, the proposed scheme deploys the trained models on an embedded edge-computing device near the inverter to reduce the communication delay. Experiments on a real embedded device achieve matched results as on the desktop PC, with about 0.1ms time cost for each inference input.