基于人工鱼群算法的应急疏散模型及优化研究

发布时间：2018-05-17 14:00

本文选题：人工鱼群算法 + 应急疏散　；参考：《湖北工业大学》2017年硕士论文

【摘要】：随着国民经济的繁荣,大型公共场所越来越密集化和规模化。公共娱乐场所具有公共性、结构复杂性、内部财产高度集中等特点,导致各种人员伤亡事故正呈不断上升的趋势。因此,为了保障建筑物内人员生命安全,我们需要研究疏散者的行为规律和影响疏散效率的各种因素等,为制定科学有效的人员疏散方案提供重要依据,有着重大的现实意义。本文以武汉沌口体育场内人员疏散方案为研究对象,主要围绕人工鱼群算法疏散模型及其优化展开了研究,主要工作如下:(1)本文利用人工智能领域中的仿生学群智能优化算法—人工鱼群算法(AFSA)作为人员疏散模型的基础建模方式,将宏观微观相结合,把疏散个体看做人工鱼智能体,定义其离散视野、步长等,并用鱼的觅食、追尾、聚群等行为模拟疏散个体的心理活动、路径选择、行为倾向等,对人员疏散过程进行仿真,最后形成整体疏散路径方案。(2)为了模拟更加细节化更加贴近于实际的疏散过程,同时也为了避免盲目疏散,本文对鱼群算法进行完善,为人工鱼定义新的行为——等待行为,构建更为科学的疏散网络,建立疏散场所三维逻辑网络,加入层次化路径引导策略,解决了以往以节点和出口直线距离作为判断依据的不现实性;并用更精确的数学表达式来刻画真实疏散环境以及人的行为和运动过程,对拥挤度、逆行、同层移动、堵塞耗时、等待时间等因素对疏散速度和路径选择的影响进行定义并均衡考虑,以疏散用时间作为目标函数提高了疏散效率。(3)人工鱼群算法疏散模型迭代完毕是将所有人疏散清空,得到一个较优解,本文令所有人工鱼回到起点进行多次循环,在每次循环结束后引入蚁群信息素,通过拥挤度繁忙度以及疏散耗时等因素综合影响的更新策略在每条边上释放一定量的信息素,改进公共板,使其在下一次循环过程中的决策受到信息素影响。在不降低疏散效率的同时提高了疏散资源的利用率,分摊出口负荷,降低路径的繁忙程度。
[Abstract]:With the prosperity of national economy, large-scale public places become more and more intensive and scale. Public entertainment places are characterized by publicity, complexity of structure and high concentration of internal property, which leads to a rising trend of casualties. Therefore, in order to ensure the safety of people in buildings, we need to study the behavior of evacuees and the factors that affect the evacuation efficiency, which provides an important basis for the formulation of scientific and effective evacuation plan, which has great practical significance. In this paper, the evacuation scheme of Wuhan Zhankou Stadium is taken as the research object, and the evacuation model of artificial fish swarm algorithm and its optimization are studied. The main work is as follows: 1) in this paper, the artificial fish swarm optimization algorithm (AFSAA) is used as the basic modeling method of the evacuation model, and the evacuation individual is regarded as the artificial fish agent. The discrete field of vision, step size and so on are defined, and the psychological activities, path selection, behavior tendency of evacuees are simulated by fish foraging, rear-end, clustering and so on, and the evacuation process is simulated. Finally, in order to simulate the actual evacuation process in more detail, and to avoid the blind evacuation, this paper improves the fish swarm algorithm and defines a new behavior-waiting behavior for artificial fish. Constructing a more scientific evacuation network, establishing a three-dimensional logical network of evacuation sites, and adding a hierarchical path guidance strategy to solve the problem of using the distance between node and exit straight line as the basis of judgment. And more accurate mathematical expressions are used to describe the real evacuation environment and the behavior and movement of people. The factors such as waiting time define and consider the influence of evacuation speed and path selection. The evacuation efficiency is improved by using evacuation time as objective function. An optimal solution is obtained. In this paper, all workers go back to the starting point for several cycles, and the ant colony pheromone is introduced at the end of each cycle. Through the update strategy which is influenced by crowded degree of busy and evacuation time, a certain amount of pheromone is released on each edge, and the common board is improved so that its decision in the next cycle is affected by pheromone. The efficiency of evacuation is not reduced, and the utilization of evacuation resources is improved, the load of exit is shared, and the busy degree of route is reduced.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】