Smart grid design modeling and prototyping

Smart grid is a promising solution for the global warming and shortage of energy, which are serious issues that the world is facing currently. A smart grid is composed of numerous micro grids. Each micro grid allows the integration of distributed renewable energy resources into the conventional electricity distribution power grid such that the goals of reduction in power cost and in environment pollution can be met through an intelligent and efficient matching between power generators and power loads. However, it is difficult for renewable energy resources to provide a reliable power supply between micro grids. Thus, renewable energy cannot be fairly distributed to all micro grids. To solve this issue, this talk proposes a multi-agent system based smart grid framework, which not only provides a fair energy allocation mechanism for micro grids, but also provides a simulation environment for actual smart grid operations. We used the Java Agent DEvelopment Framework (JADE) to simulate multi-agent based micro grids, and the Matlab/Simulink simulation platform to design and implement power generators and power loads. Model-predictive control (MPC) method is also proposed for smart grids. Experiment results show that the proposed method is fairer than other state-of-the-art methods and MPC related issues are also resolved.