AGENT-BASED MODELING OF URBAN RECOVERY: CONSTRUCTING AN UMBRELLA OF RECOVERY PLANS VIA VIRTUAL CITY TWIN AND ITERATIVE ALGORITHM APPROACH.

This work presents a generalized iterative algorithm of short-term recovery management after large-scale disasters. The proposed method enables the coordination of reconstruction processes when the initial information is fluid and uncertain. The demonstrated algorithm supplies the recovery administration with a set of generated recovery plans, which may be compared from multiple perspectives. The city is modeled adopting the PCANS framework (Krackhardt et al., 1988), enabling to simulate of the complex interactions between individual recovery agents and recovery teams, recovery teams and individual infrastructure links, and between individual infrastructure links and the performance of a particular infrastructure or the whole city. The resulting model forming a virtual twin of an urban area is analyzed numerically with a set of different restoration strategies and comparing the efficiency of the resulting plans. As a result, the recovery coordinator is provided with a set of optional restoration strategies, which enables avoidance of the delay of recovery initiation caused by the ambiguity of the initial state of the system and the impossibility of defining an optimal recovery plan during the short-term recovery phase. [ABSTRACT FROM AUTHOR]