Complex network theory-based optimization for enhancing resilience of large-scale multi-energy System11The short version of the paper was presented at CUE2023. This paper is a substantial extension of the short version of the conference paper

To sustain energy supply capability and mitigate vulnerability to operational failures, enhancing resilience in multi-energy systems (MES) becomes imperative in the face of deliberate attacks and natural disruptions. In this paper, a resilience enhancement optimization model (REOM) is proposed, incorporating various enhancement measures such as adding backup, strengthening physical protection, and constructing new lines. The objective is to minimize the betweenness centrality loss of the MES after being attacked or disrupted, where the betweenness centrality is improved by integrating the power output of energy supply equipment as weighted factors. To accelerate the calculation of shortest paths in large-scale MES, an approximate algorithm for shortest paths is proposed to extract necessary information of shortest paths, through selecting landmarks and converting the target path into the summation of different segments with the landmarks. Then, the REOM is transformed into a bi-level problem while the lower-level one is linearized into a mixed-integer linear programming (MILP) problem, and a genetic algorithm-MILP (GA-MILP) is employed to solve it. Assessing the optimal resilience enhancement plan, a comparison is made with the results obtained under two different attack strategies in the remaining two scenarios through three validation indicators: network fragmentation, decline in energy transfer efficiency, and load loss rate. Through simulations on both a 67-bus power-gas system and a virtual European power-gas system, it was observed that the weighted betweenness centrality loss decreased by 45.37% and 14.95%, respectively, compared to scenarios without the incorporation of the optimal resilience enhancement plan, effectively reinforcing system robustness.