Dynamic Response Characteristics Analysis of Energy Storage System Based on Multimodal Data Fusion and Methods for Improving Power Supply Stability in Data Centers

Résumé fourni par la source

With the rapid development of the digital economy, the scale of data center computing power continues to expand, and the high-density, high load, and uninterrupted power supply demand puts forward extreme requirements for the dynamic response capability and power supply guarantee accuracy of the supporting energy storage system. The traditional data center energy storage power supply control mode relies on a single electrical monitoring data, which has defects such as one-sided state perception, delayed dynamic response, weak disturbance suppression ability, and poor adaptability to multiple working conditions. It is difficult to cope with power supply instability caused by complex working conditions such as grid voltage fluctuations, instantaneous load changes, aging of energy storage modules, and environmental parameter drift. In response to the above pain points, this article proposes a dynamic response characteristic analysis method for energy storage systems based on multimodal data fusion and an optimization system for data center power supply stability.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.