A cluster-based emergency vehicle routing problem in disaster with reliability

In the event of natural disasters, relief distribution is the most challenging problem of emergency transportation. What is important in response to disaster is victims’ relief in disaster areas with the quick distribution of vital commodity. In this regard, damage to infrastructure (e.g., roads) can make trouble in designing a distribution network. So, this paper considers a problem using a threestage approach. In the first stage, preprocessing of model inputs is done through an artificial neural fuzzy inference system (ANFIS) followed by investigating the safest route for each cluster using of decisionmaking techniques and graph theory. In the second stage, a heterogeneous multidepots multimode vehicle routing problem is formulated for minimizing the transportation time and maximize the reliability. Finally, since the routing problem is NPhard, two multiobjective metaheuristic algorithms, namely nondominated sorting genetic algorithm (NSGAII) and multiobjective firefly algorithm (MOFA), are proposed to obtain the optimal solution and compared their performance through a set of randomly generated test problems. The results show that for this routing problem, the MOFF gives better solutions in comparison to NSGAII and performs well in terms of accuracy and solution time.