基于关联学习的配电网多年投资规划模型

Multi-year Investment Planning Model of Distribution Network Based on Correlation Learning

为保障配电网可靠性投资效益的最大化,研究配电网精确投资规划技术尤为重要。当前大部分的配电网投资规划研究主要是在计及完备潮流方程与网络安全约束的基础上建立模型,在面对大量投资措施决策时,上述模型在计算量与收敛性方面都面临挑战。针对上述模型的问题,提出了一种基于投资措施与配电网可靠性关联学习的配电网多年投资规划模型,采用两者关联性取代复杂的潮流与网络安全约束,进而计算每种投资措施对可靠性指标的提升效果,考虑投资措施的多年约束,获得最优方案组合,并通过实际电网对所提投资规划模型的有效性进行验证。

In order to ensure the maximization of investment benefit such as power supply reliability,researches on the accurate investment planning technology is particularly important.At present,most of the research on distribution network investment planning mainly establish investment planning optimization model based on complete power flow equation and network security constraints.However,in the face of heavy investment decision,the traditional model is faced with challenges in computation and convergence.Therefore,in the light of the above problem,a multi-year investment planning model is putforward based on the correlation between investment strategy and reliability of the distribution network.In the model,the correlation constraint is adopted to replace the complicated power flow and network security constraints.Considering the multi-year constraints of the investment measures,we calculate the improvement effects of various reconstruction strategies on the reliability index,and then the optimal investment plans of distribution network can be obtained.The validity of the proposed investment planning model is verified by the actual grid.