基于蒙特卡洛算法的配电网规建运供电可靠性分析

Parameter Reliability Analysis of Distribution Network Planning and Transportation Automatic Control System based on Monte Carlo Algorithm

研究分析一种建立在配电系统基础上的微电网运行可靠性管理的数字平台。首先,介绍了配电网可靠性管理数字系统的概念和运行原理。然后,为确保该数字系统方案主要解决可靠性和经济性两方面问题,提出了管理配电网可靠性数字系统的基本原则。最后,提出了基于机器学习、蒙特卡洛方法的两种评估配电网可靠性的算法,对于评估系统可靠性具有通用性。第一种算法利用机器学习计算微系统状态的运行参数,通过训练模型确定可靠性指标。第二种算法直接使用机器学习方法分析系统可靠性指标,通过对模型的训练,立即确定配电网系统的可靠性指标。通过在IEEE 96总线RTS实验中应用上述算法,结果显示,在确保微系统可靠性评估准确性的条件下,所提出的两种算法速度分别为110 s与2 s,相较于传统算法的184 s,具有更高的评估速度。所以,提出的基于机器学习方法的两种系统可靠性评估算法能兼顾精度和速度地确定可靠性指标,从而评估和确保配电网的供电可靠性。

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 reliability are introduced.Then,in order to ensure that this digital system solution mainly solves the problems of both reliability and economy,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 indicators using machine learning methods and immediately determines the reliability indicators of the distribution network system by training the model.By applying the above algorithms in the IEEE 96-bus RTS experiments,the results show that the speed of the two proposed algorithms is 110 s and 2 s,respectively,which has a higher evaluation speed compared with the 184 s of the traditional algorithms,under the condition of ensuring the accuracy of the microsystem reliability evaluation.Therefore,the two system reliability assessment algorithms proposed in this paper based on the machine learning method can determine the reliability indexes with both accuracy and speed,so as to assess and ensure the power supply reliability of the distribution network.