Algorithm Research On Single-source Emergency Evacuation With Hazardous Chemicals

With the continuous development of science and technology, chemicals are widely used in daily production and life. Recently, with the highly de-velopment of technology and the frequency of extreme weather, the harm of hazardous chemicals accidents become bigger and bigger. If we can response these hazardous chemicals accidents in a scientific and rational way, we can reduce people’s loss for both life and property. we can also reduce the negative impact of society. Therefore, to reduce the hurt people get in the accidents as much as possible, this paper will study the emergency evacuation problem of hazardous chemicals.firstly, this paper summarizes the development of four kinds of emergency evacuation model. In former researches about emergency evacuation problems, researchers usually assume that there is no wait in the process of evacuation. Considering that the emergency evacuation of hazardous chemicals not only needs the speed as fast as possible, but also needs the hurt people get as little as possible. Therefore, this paper proposes an idea that people can wait in the internal nodes during the evacuation. Then, this paper proposes a waited emergency evacuation arithmetic of hazardous chemicals accident, in which situation people can get less hurt than they get in no-wait situation during the evacuation.Secondly, this paper summarizes three kinds of leakage model of hazardous chemicals and two kinds of diffusion model of hazardous chemicals. Based on these models, this paper proposed a new method to calculate the accumulat-ed concentration. Compared with the previous calculation method, the new method closer to the real situation. It is concluded that the emergency evacu-ation project based on the new method is more meaningful. This paper verifies the conclusion by a simulation.Finally, to solve no-wait and waited emergency evacuation arithmetic of hazardous chemicals accident problem, this paper proposes three kinds of heuristic algorithm, and implement the algorithm with a simulation. Then, to verifies the accuracy and efficiency of these algorithms, this paper compares conclusions between these algorithms, compares conclusions with the optimal solution, and compares the computing time between simulations with different scale.