BOTTLENECK VEHICLE ROUTING FOR PLANNING POWER SUPPLY DURING DISASTERS

In the event of a widespread power outage due to a disaster, telecom companies will run a backup power supply in the telecom buildings to maintain services. Since the amount of reserve power decreases with time, each building needs to replenish power from a power supply vehicle. To minimize the impact on telecommunication services during power outages, the time of power failure in telecommunication buildings must be minimized. The optimal route of the power supply vehicle is the one that ensures the shortest possible outage time at the affected buildings. Since delays must be compared per building, the problem of searching for this route is difficult to represent as a standard vehicle routing problem. Instead, this paper introduces the Bottleneck Vehicle Routing Problem, which asks for a route plan that minimizes the maximum damage caused by the outage of power. A branch-and-bound method for the single-vehicle case and a column generation method for the multiple-vehicle case are presented. These methods are shown to be effective through numerical experiments.