文摘
An optimization model is developed to guide recovery of a disrupted water distribution system. The model minimizes the total cost of recovery, including the disruption cost of unmet demand during the repair process and the repair cost itself. The optimization schedules repair tasks under precedence and resource constraints and contains an embedded flow problem that optimizes the distribution of water in each time period, given the state of the network. A simulated annealing algorithm is developed for scheduling the tasks, with the embedded flow problem solved using a generalized reduced gradient method. Experiments with a test water distribution system confirm the effectiveness of the model and provide insight regarding the effects of limited resources available for recovery and of the usefulness of having multiple modes for execution of specific tasks.