LEARNING-AIDED BIGRAPH MATCHING APPROACH TO MULTI-CREW RESTORATION OF DAMAGED POWER NETWORKS COUPLED WITH ROAD TRANSPORTATION NETWORKS

The resilience of critical infrastructure networks (CINs) after disruptions, such as those caused by natural hazards, depends on both the speed of restoration and the extent to which operational functionality can be regained. Allocating resources for restoration, a combinatorial optimal planning problem that involves determining which crews will repair specific network nodes and in what order, is complicated by several factors: the connectivity between the CIN and the road transportation network used for travel by repair crews, the enormous scale and nonlinear behavior of these networks, and the uncertainty in repair times. This paper presents a novel graph-based formulation that merges two interconnected graphs, representing crew and transportation nodes and power grid nodes, into a single heterogeneous graph. To enable efficient planning, graph reinforcement learning (GRL) is integrated with bigraph matching. GRL is utilized to design the incentive function for assigning crews to repair tasks based on the graph abstracted state of the environment, ensuring generalization across damage scenarios. Two learning techniques are employed: a graph neural network trained using Proximal Policy Optimization and another trained via Neuroevolution. The learned incentive functions inform a bipartite graph that links crews to repair tasks, enabling weighted maximum matching for crew-to-task allocations. An efficient simulation environment that pre-computes optimal node-to-node path plans is used to train the proposed restoration planning methods. An IEEE 8500-bus power distribution test network coupled with a 21 sq km transportation network is used as the case study, with scenarios varying in terms of numbers of damaged nodes, depots and crews. Results demonstrate the approach's generalizability and scalability across scenarios, with learned policies providing 3-fold better performance than random policies, while also outperforming optimization-based solutions in both computation time (by several orders of magnitude) and power restored.