Energy Interconnection Network Architecture Design Based On SDN And Research Of The Control Domain

Energy internet is very important in solving the problem of energy and environment, so scholars have launched research work actively, and the design of the energy internet architecture is very important for the development of energy internet. This paper presents an general architecture of the energy network, and propose an algorithm for the control domain allocation problem of control layer in the architecture.This thesis on basic of a concrete analysis of energy interconnection network architecture of the current several mainstream models, and with reference to some research results at home, puts forward an overall design idea of energy Internet architecture based on software defined network technology. It concretes describes the important functional components from the bottom of information collection to the upper of control layer and their connection mode, such as the bottom using a new device (called Intelligent Energy Controller) to collect and upload bottom information and execution control command from the upper layer, and in control layer data centers will have backup controller in order to increase the system stability and reliability. In addition, the paper also describes in detail the whole run process of system, and the structure characteristics are analyzed, pointing out the advantages of the architecture and the areas that need further research and improvement.Then put forward an equal allocation algorithm for problem of the controllers how to allocate data centers, using Visual C++ 6 to realize the algorithm, and using MATLAB 7 to draw analyze the results. Use the algorithm to initialize the system and find that the algorithm can be very good in most cases for allocating data centers for controllers reasonably, but there are special circumstances need further analysis and research. Increase the data center node of one controller, and calculate the whole network and part respectively, will find that when increase a few nodes, the two algorithms is the same, and when the number of larger, better of the whole network.