| 英文摘要 | In recent years, public emergencies have occurred frequently and caused huge lost of human production and life. Transportation systems play an important role in emergency response. Without active and effective management of transportation systems, we will not be able to quickly respond to public emergencies. As human behavior under public emergencies has great impact on emergency transportation management, this dissertation studies public emergency’s impact on transportation systems and corresponding strategies from the perspective of human behavior. This dissertation is based on the ACP (Artificial Systems, Computational Experiments, Parallel Execution) theory which is a new theory on modeling, analyzing, managing and controlling complex systems. In this dissertation we first study the artificial population generation method and human traffic behavior under public emergencies. Then, a prototype of artificial transportation systems for emergency (ATS-E) is developed. The impact of human behavior on transportation systems under public emergencies are studied based on the ATS-E and computational experiment design theory, and some emergency transportation management policies are proposed from the standpoint of human behavior intervention. The main work of this dissertation is as follows. Firstly, we introduce definitions of public emergency and give a comprehensive literature review on human behavior under public emergencies, emergency evacuation modeling and emergency evacuation control. Secondly, we present our study on artificial population generation. As the traditional population synthesis methods are not suitable under the condition that population statistical characteristics change, this dissertation proposes a new artificial population generation algorithm. The algorithm is based on genetic and evolutionary concepts, and contains four steps, namely, determining artificial population properties, coding artificial population properties, generating initial artificial population at random and updating artificial population in accordance with genetic and evolutionary rules. Thirdly, a computational platform specifically for emergency, ATS-E, is developed based on TransWorld, an existing artificial transportation system prototype developed in our lab. The ATS-E takes into account the representation of public emergencies, human behavior and background traffic flow generation. Fourthly, in order to obtain valid and objective results, this dissertation st... |
修改评论