高温天气下配电网综合韧性评估方法

Research on Comprehensive Resilience Assessment Method of Distribution Network under High Temperature Weather

随着极端天气导致的自然灾害频发,针对配电网的综合韧性评估难以采用传统可靠性评估方法,提出一种高温天气下配电网综合韧性评估方法。首先,利用核岭回归(kernel ridge regression, KRR)方法,提出基于高温天气片段的配电网节点负荷率估计方法。其次,针对高温天气对配电网产生的影响,建立了高温引起潮流变化导致配电网故障的概率模型,同时考虑了高温对配电网元件的影响,利用脆弱性曲线方法建立了配电网高温天气下元件故障概率模型,进而提出高温天气导致的配电网综合故障概率模型。再次,从捕捉配电网韧性功能曲线特征角度出发,提出配电网综合评估指标模型,并利用泊松分布方法获取配电网韧性场景。最后,在北京某地区双环网配电网系统算例中验证了所提方法和指标的有效性,通过与传统韧性评估指标对比表明了所提配电网综合韧性评估的优越性。

With the frequent occurrence of extreme weather leading to natural disasters,it is difficult to use traditional reliability assessment methods for the comprehensive resilience assessment of distribution networks.This paper proposed a comprehensive resilience assessment method for distribution networks under high temperature weather.First,using the kernel ridge regression(KRR)method,a distribution network node load rate estimation method based on high-temperature weather segments was proposed.Secondly,in view of the impact of high temperature weather on the distribution network,a probability model of faults caused by changes in power flow caused by high temperature was established.At the same time,the impact of high temperature on the components of the distribution network was considered,and the fragility curve method was used to establish a fault model for distribution network components.And then proposed a comprehensive failure probability model of distribution network caused by high temperature weather.Thirdly,from the perspective of capturing the characteristics of the distribution network's resilience function curve,a comprehensive evaluation index model AR for the distribution network was constructed,and Poisson distribution method to obtain distribution network resilience scenarios was used.Finally,the effectiveness of the proposed method and indicators was verified in a double-ring distribution network system example in a certain area in Beijing.The superiority of the comprehensive resilience assessment of the distribution network proposed in this paper was demonstrated through a comparative study with traditional resilience assessment indicators.