Deterministic heuristics in railway disaster relief planning

During a natural disaster, timely relief activities can save human lives. Effective planning of such relief operations often necessitates solving complex optimization problems. Stochastic algorithms are widely used to generate solutions within limited computation time. However, there is no guarantee that the solution is or even close to the global optimum. Worse still, the quality of the solution, i.e., the effectiveness of the disaster relief planning, is stochastic. In this work, we study the problem of dispatching commodities to the disaster area by rail and propose a number of polynomial-time heuristics to generate a deterministic solution. A small case study, which can be exhaustively searched, shows that the global optimum can be achieved. In practice, these deterministic heuristics can be deployed in parallel with stochastic algorithms to ensure effective disaster relief operations.