|
As the development of modern technology, both categories and quantities of the hazardous chemicals increased tremendously, but the chemicals related accidents, especially release ones, happen all the times during their production, storage and usage, as is greatly threatening the health of residents and the sustainable development of environment. For example, a gas leak in Chongqing, China in December 2003 spread toxic hydrogen sulphide across mountain villages, killing 243 people in one of China’s deadliest industrial accidents. More than 41,000 villagers were forced to evacuate from their homes and thousands of survivors suffered lung damage and burns on their eyes and skins. More recently, in 2007, over 113,000 chemical release accidents happened in China, and about 229,000 residents was rescued or evacuated. Also in the United States, there have been,32,022 chemical related spill events reported in 2007. Therefore, it is badly necessary to explore how to make good use of available emergency resources to establish corresponding evacuation scheduling model and its algorithms, aiming at reducing the loss caused by chemical release accidents.The thesis analyzes the characteristics of emergency evacuation, basic problems studied by other researchers, and we categorize the evacuation problem from several aspects. Also, we present a state-of-art to both the evacuation planning and corresponding policies from the macroscopic level, and modelling methods and relevant algorithms from the microscopic level. Then the thesis studies the evacuation problem in the scenario of chemical release, which is divided into two sub-problems according to whether the evacuees hold the ability of self-evacuating, that is, active evacuation problem and passive evacuation problem:(1) Regarding to the passive evacuation problem, this thesis divide it into two stages to reduce the complexity, that is, how to dispatch the evacuation vehicles to affected areas, then how to transfer the evacuees to safe areas as soon as possible. We model the problem with the grouped, staged, and prioritized strategy, then develop an effective in-time greedy rule based heuristic algorithm:vehicle scheduling and routing to passive evacuation algorithm (VSR-PE Algorithm). (2) To active evacuation problem, our research mainly focuses how to evacuate the residents in the affected area to safe area as soon as possible. At the same time, different from passive evacuation problem, traffic capacity is considered as a constraint in our model. Based on the analysis of the optimality of active evacuation problem from the perspective of theory, we develop a heuristic algorithm:traffic capacity constrained routing algorithm to active evacuation (TCCR-AE Algorithm)Finally, simulation experiments are designed to validate the effectiveness and their scalability to large scale problem of VSR-PE and TCCR-AE algorithm.
|