考虑期望损失的综合能源系统故障智能筛选和排序方法

综合能源系统实现了电、热、天然气等多种能源系统的耦合互联,相比于单一能源系统,其故障状态数目大幅增加,如何快速准确地从系统海量的故障状态中筛选出最严重的故障,从而提高风险评估效率,成为综合能源系统风险评估中的关键难题。针对这一问题,基于遗传算法提出了一种综合能源系统故障智能筛选和排序方法。首先,计及多种能源网络的约束条件,建立了综合能源系统故障期望损失双层优化模型,同时刻画了故障发生概率和故障损失对系统的影响。进而,基于遗传算法提出了故障期望损失双层模型的求解方法,并对传统遗传算法进行改进,使得方法能够同时筛选出期望损失最严重的多个故障状态。最后以修改的IEEE 33节点电力系统、巴厘岛供热系统和比利时天然气系统组成的综合能源系统为例,仿真结果表明所提方法的故障筛选效率明显优于常规枚举法和蒙特卡洛方法,证明了所提方法的有效性。

Contingency Intelligent Screening and Ranking Approach for Integrated Energy System Considering Expected Loss

The integrated energy system(IES)realizes the coupling and interaction of various energy systems such as electricity, heat and natural gas. The contingency number of IES increases dramatically compared with the traditional simplex energy system. It has become a concern for the risk assessment of IES how to screen the most serious contingencies from the mass number of contingencies fast and accurately, which is important for improving the efficiency of risk assessment. Focusing on this topic, a contingency intelligent screening and ranking approach for IES is proposed based on genetic algorithm. Firstly, considering the constraints of various energy networks in IES, a bi-level optimization model is established to depict the expected loss of IES caused by contingencies, which reflects the influences of contingency probability and economic loss caused by contingency. Furthermore, a solution algorithm of the bi-level model is proposed based on genetic algorithm. With the improvement of traditional genetic algorithm, the proposed solution algorithm can simultaneously screen multiple contingencies with highest expected loss. Finally, the case study is based on an IES test case which consists of the modified IEEE 33-bus electric system, Barry Island heating system and Belgian gas network. Numerical results indicate that the computational efficiency of proposed approach is obviously higher than that of traditional enumeration method and Monte Carlo method, which verifies the effectiveness of the proposed approach.