Cooperative Control Of Micro-grids Based On Multi-agent Technology

In recent years, with the rapid development of the global economy and the improvement of people's living standards, energy crisis and environmental problems have become a major limit the development of human beings. In order to deal with these problems, the distributed generation technology has emerged. However, the access of numerous distributed generations(DGs) will bring severe impact to the main grid. As an important form of distributed generation, micro-grid can fully utilize the advantage of distributed generation, so as to eliminate the bad influences. There are two kinds of operation modes, namely, the grid-connected mode and the islanded mode. Normally, the micro-grid connects with the main grid to supply power to the loads together. Whe n the main grid is disturbed or because of some necessary reasons, the micro-grids will disconnect with the main grid to islanded mode. The primary control and the secondary control are used to maintain the stability of the voltage and frequency. At this time, the micro-grids supply power to the local loads alone. Another important goal of the micro-grids is to achieve economic operation and optimal scheduling. The traditional secondary control and economic dispatch are centralized, which has a huge amount of communication to reduce the security and reliability of the system. The multi-agent system and distributed technology are introduced against these problems. The main contents of this paper are as follows:(1) A distributed cooperative control structure for micro-grids is studied instead of the traditional centralized structure. This structure consists of the primary controller, the distributed secondary control device and the distributed economic dispatching module. The control structure does not need a central controller and that make the communication network sparse relatively. It should be noted that communication structure can be simple because of the desired plug-and-play capability of micro-grids. Therefore, this structure greatly improves the scalability and robustness of the system.(2) An adaptive distributed secondary cooperative control strategy is fully designed. Different from the conventional secondary control algorithm, this strategy is distributed such that each generation unit can only get information from its own bus node and the neighbor bus. At the same time, the neural network is used to estimate the nonlinear part of the system to the synchronous tracking in finite time.(3) A distributed algorithm is proposed to solve the economical dispatch problem. Based on multi-agent system theory and distributed technology, the controller only needs to know local information but not the global information. The proposed approach can minimize the total cost of generation without the power generation capacity constraints and with it.