Distributed genetic algorithm for disaster relief planning

The problem studied in this paper is the management of vehicle routing in case of emergency. It is decomposed into two parts. The first one deals with the emergency planning in the event of receiving a set of requests for help after a major disaster such as in the case of an earthquake, hurricane, flood, etc. The second part concerns the treatment of contingency as the arrival of a new request or the appearance of a disturbance such as breakdowns of vehicles, the malfunction of roads, availability of airports, etc. To solve this problem we proposed a multi-agents approach using a guided genetic algorithm for scheduling vehicle routing and local search for the management of contingencies. The main objectif of our approach was to maximizing the number of saved people and minimizing the costs of the rescue operation. This approach was tested with the modified Solomon benchmarks and gave good results