Machine Learning based and Multi-Agent System based Control and Optimization Approaches for Electric Vehicles, Power Grids, and their Interactions

The purposes of this research are to stabilize the balance between demand and supply of electricity in the energy grid, to improve the system ability to cope with sudden changes in electricity demand and supply, and to optimize performance and efficiency of storage systems, e.g., batteries, expected to be used in home and regions.

As a methodology to realize them, Prof. Nguyen proposes distributed implementation of the solutions, such as battery State of Charge (SoC) monitoring, charge/discharge control, and machine-learning based demand and supply prediction, onto Multi-Agent System (MAS). In this system, a wide range of power management functions required from in-vehicle battery to home, region, and even inter-region are executed by one or multiple agents at different scales. Its way of system design and operation is consistent, which is well advanced and unique in this domain. Prof.

Nguyen's research shows practical paths to solve the problems of upcoming energy network system such as renewable energy integration and emergency power management.