期刊名称：International Journal of Grid and Distributed Computing
印刷版ISSN：2005-4262
出版年度：2016
卷号：9
期号：10
页码：347-356
出版社：SERSC
摘要：It is one of the effective methods to improve the reliability of the distribution system by using looped network power supply. With the development of new energy technology, distributed generation has been more widely used. But a small power system named micro-grid which is assembled by the device such as distributed generation, energy storage device and controllable load in a certain operating rules catches the attention of people. It can operate in islanded mode with the power grid failure, which can greatly improve the reliability of the system. Traditional analytical method and Monte Carlo method are difficult to evaluate its reliability accurately. In this paper, a new artificial intelligence method based on Bayesian networks is used for the system reliability analysis and evaluation. The corresponding Bayesian networks are established for the reliability test bus 2 which containing looped network and the improve reliability test bus 6 which containing micro-grids separately. With the help of causal reasoning and diagnostic reasoning, the supply influence of different types of DG output intermittent, operation mode on loads are detected, the key factors which have an impact on important load power supply reliability are found out, which is the theoretical foundation for practical engineering decision, so as to improve system reliability, identify failures and make maintenance scheduling.
关键词：micro-grid;looped network; distribution system; reliability evaluation; ;Bayesian network
