Provincial Electricity Authority (PEA) & Asian Institute of Technology (AIT)
This research project, conducted jointly by the Provincial Electricity Authority (PEA) and the Asian Institute of Technology (AIT), developed an AI-Based Smart Microgrid Platform System aimed at increasing the stability and efficiency of electricity generation and distribution. The project designed and built a scaled Smart Microgrid experimental set modelled on the real-world distribution network of Mae Sariang District, Mae Hong Son Province, incorporating energy sources including solar photovoltaic (PV) panels, a hydroelectric generator, a diesel generator, and battery storage systems.
At the core of the system is an AI-driven control architecture combining a Weather Research and Forecasting (WRF) model with a Temporal Fusion Transformer (TFT) for multi-feeder power demand prediction, and a Particle Swarm Optimization (PSO) algorithm for energy dispatch planning. These components work together to forecast electricity demand, optimize battery charge/discharge schedules, and reduce peak power consumption. A low-level microgrid controller translates these optimization plans into real-time setpoints for physical devices, with communication handled via an IEC61850 protocol layer implemented on Raspberry Pi units connected through Modbus RTU to TCP/IP converters.
The platform supports multiple operational modes: Grid Connect, Intentional Islanding, Unplanned Islanding, and Outage recovery. It also includes automated transition logic ensuring stable and seamless switching between states. A web-based UI built on FastAPI allows operators to monitor device status, view simulation results, and issue control commands. Despite procurement delays caused by COVID-19 supply chain disruptions, the experimental kit was successfully assembled and tested at the Distributed Education Center (DEC) building, with final installation planned at the High Voltage Training Center in Nakhon Pathom Province. The project demonstrates a viable, scalable framework for intelligent microgrid management applicable to remote and grid-edge communities in Thailand.
Fig 1: Single Line Diagram of the Micro Grid System
Fig 2: Day-ahead Scheduling of the Micro Grid System
Fig 3: Main Power Control Cabinets of the Smart Micro Grid System
Fig 4: Information transmission architecture within the Smart Microgrid experiment set
Fig 5: Training on the Smart Microgrid
Fig 6: Training on the Smart Microgrid
