Next-Gen Science Park VPP and Microgrid Architecture: NCKU Proposes AI Energy Management Framework
MEP · Smart operations · AEC industry
A National Cheng Kung University research team proposed an integrated energy management framework for next-generation science parks, addressing high-tech power quality demands and 24/7 CFE mandates. The scheme integrates hybrid AC/DC microgrids, AI dual-track rolling scheduling, and blockchain VPPs to enable peak shaving, interface with Taipower ancillary services, and provide black-start and reverse grid support capabilities.
A technical article published in CECI Engineering Technology by an NCKU electrical engineering research team (Jian-Tang Liao, Yu-Cheng Chen, Yu-Cheng Yen, and Hong-Tzer Yang) examines AI-driven energy management for next-generation science park virtual power plants and microgrids. Addressing semiconductor sensitivity to voltage sags and high-voltage DC data center needs, the paper presents a three-tier architecture comprising AI EMS, blockchain VPP trading, and DERs to transform science parks into resilient prosumers.
Key takeaways
- The AI EMS implements dual-track scheduling combining day-ahead planning with rolling horizon control every 5 to 15 minutes to counter renewable intermittency and optimize peak demand charges.
- The platform leverages blockchain distributed ledgers and smart contracts for 24/7 CFE matching and transparent energy trading, interfacing with Taipower ancillary services via IEC 61850 and OpenADR protocols.
- The physical infrastructure deploys hybrid AC/DC microgrids with solid-state transformers connecting AC grid to DC 800V buses, integrating SOFCs, storage, and IT racks for grid-forming, black-start, and external island support.
Source：中華顧問工程司／台灣世曦｜中華技術 Read the original ↗
Prepared, checked and automatically published by AI; not individually reviewed by a human. Translation checked by AI.