Research On The Coordination Control Strategy Of Active Distributed Networks Based On Multi-agent Technology

Compared with centralized generation,distributed generation mainly including solar power, wind power, small scale hydro power, biomass power and so on is a new way of generating electricity. Although it is rapidly developed as an ideal way of utilizing renewable energy source due to its low cost, high reliability, cleanliness and flexibility, its negative effects on the electricity generation become more and more obvious. Since distributed generation mostly apply new energy in power generation, it is unpredictable, uneasy to control, and unstable, which make the power flow in traditional power distribution network change its direction in accordance with the alteration of distributed generation. This alteration would cause quality reduction of the whole distribution network and even incur power flow congestion. These problems lead to a relatively low composition ratio for distributed generators within the whole power system, which actually limits the generation capacity of distributed generation and makes it unable to merge into power grid causing waste of resources. Active distribution networks can realize the active management for the controllable factors in power distribution network, which can effectively handle the problems due to the insertion of distributed generators in distribution network. Focusing on the problems above, this thesis builds coordination control system based on multi-Agent technology, and analyzes the power flow management of active distribution networks method under normal condition and the power recovery method of distribution network when the main power net breaks down.First, the thesis introduces the research status of active distribution networks home and abroad, and gives an analysis of its structure and coordinated control system. Then it analyzes the multi-agent technology and puts forward its application in active distribution networks. According to the superincumbent layered structure, it builds a coordinated control system for active distribution networks based on multi-Agent technology.Secondly, the thesis analyzes the power flow management in active distribution networks, building a mathematical model for it, and transferring it into the matter of generating the max current with the minimum cost in graph theory. Based on the basic theory of graph theory and modeling approach, the thesis builds the Graph theory model.Thirdly, the thesis makes a thorough analysis of the classic algorithm in graph theory. Based on the algorithm and combine it with the features of distribution networks graph theory model,the thesis further modifies the algorithm of maximizing current with the minimum cost, uses it to figure up the graph theory model and design a power flow management strategy for active distribution networks. The result has been proved in MATLAB’s simulation test.At last, the thesis makes an analysis for the isolate island operation of active distribution networks, and builds the mathematical model of recovering power when it is in isolate island state. The thesis puts forward a fully distributed multi-Agent system, making the power distribution network be able to remain a regional power balance and keep the power of important loads.