No abstract

No abstract

Nowadays, tsunami is becoming one of the most dangerous natural disasters. Most coastal regions face a potential risk of this kind of disaster. Along with the early warning system, evacuation is one of the first mitigation procedures to consider. The evacuation simulation then has received a lot of studies in recent years in the domain of computer science. In fact, there are always the part of evacuees (e.g. the tourist) who lack information of the evacuation map, which motivates us to study the problem of optimizing of guidance sign placement for tsunami evacuation.With regards to the problem of optimizing of guidance sign, the simplest approach is to build the simulation of evacuation and then optimize the input parameters (the roads to deploy the signs) to maximize the survivals. This approach presents the limitation of computational requirement for exploring the huge parameter space. In the recent works, the Linear Programming approach was introduced to optimize the positions and directions of signs to minimize the average evacuation time. However, this linear approach did not to address the number of survival which is considered the ultimate goal of the problem.In this paper, we first propose the approach of Linear Programming that optimizes the directions of predefined positions of signs. Then, we present an approach for optimizing the sign placement by using genetic algorithm with the fitness evaluated by the Agent-based Simulation.

Nowadays, tsunami is becoming one of the most dangerous natural disaster for coastal regions. Along with the early warning system, evacuation is one of the first mitigation procedures to consider. The evacuation simulation then received a lot of studies in recent years in the domain of computer science. In fact, there are always the part of evacuees (e.g. the tourist) who lack information of the evacuation map, which motivates us to study the problem of optimizing of guidance sign placement for tsunami evacuation. In this paper, we first propose the approach of Linear Programming that speeds up the Evaluation of Casualties in Agent-based Simulation in order to overcome the problem of computational speed of agent-based simulation. Then, we present an approach for optimizing the sign placement by using genetic algorithm with the fitness evaluated by the agent-based simulation.