The aim of this study is to investigate and analyze a digital platform for reliability management of microgrid operation based on the distribution system.First,the concept and operational principles of the digital system for managing distribution grid reli-ability are introduced.Then,in order to ensure that this digital system solution mainly solves the problems of both reliability and e-conomy,the basic principles of managing the digital system for distribution grid reliability are proposed.Finally,two algorithms for assessing distribution network reliability based on machine learning and Monte Carlo methods are proposed,which are generalizable for assessing system reliability.The first algorithm utilizes machine learning to calculate the operational parameters of the microsystem state and determines the reliability indicators by training the model.The second algorithm directly analyzes the system reliability indi-cators using machine learning methods and immediately determines the reliability indicators of the distribution network system by train-ing the model.By applying the above algorithms in the IEEE 96-bus RTS experiments,the results show that the speed of the two pro-posed algorithms is 110 s and 2 s,respectively,which has a higher evaluation speed compared with the 184 s of the traditional algo-rithms,under the condition of ensuring the accuracy of the microsystem reliability evaluation.Therefore,the two system reliability as-sessment algorithms proposed in this paper based on the machine learning method can determine the reliability indexes with both accu-racy and speed,so as to assess and ensure the power supply reliability of the distribution network.
维普
万方数据