Research On Equilibrium Model And Optimization Method For Campus Evacuation

In the case of emergency evacuation, especially on the campus, as the positions of the evacuation objects are different, the corresponding judgments and route choices of their spatial-temporal routes vary from one to another. Concerning the campus evacuation with limited resources, it is very important to reasonably optimize the assignment of spatial-temporal routes for each evacuee and improve the performance of emergency evacuation from the perspective of individual need and system efficiency. Nowadays it becomes an important research subject and urgent issue in the fields of modern traffic discipline, artificial intelligence, computer science, modern communication discipline, as well as public health and safety. For this issue, most of the traditional traffic assignment models consider from the perspective of evacuation decision makers and pursue the highest system efficiency with the loss of someone’s interest as the price, ignoring the individual need in the practical environment. This dissertation takes into account the evacuation equilibrium model concerning the balance of both individual need and system efficiency, and makes research on indoor evacuation in teaching buildings and outdoor evacuation in the road network on campus. The research content and results include the following aspects:1) Taking into account the individual need and system efficiency, this dissertation establishes an evacuation equilibrium model which can improve the practicality of evacuation planning programs. Most of the traditional route planning programs consider the evacuation planning problem from the perspective of evacuation decision makers with the shortest time or other target. But in real situation, each individual tends to look for his most favorable route, which makes the original evacuation plan loses its viability. For this case, with the idea of evolutionary algorithm and user equilibrium which means each person will choose his favorable path in the current circumstances, this dissertation establishes the evolutionary wardrop equilibrium. In the meantime, in order to assess the equilibrium degree of evacuation progress, the equilibrium index is proposed and it provides more effective and feasible evacuation plans for decision makers.2) Based on the cellular automata model, this dissertation proposes the concept of "steady" and establishes an indoor evacuation equilibrium model for teaching buildings. As a special place on campus, the teaching building has its unique characteristics:the building’s layout is relatively complex, the distribution of students is relatively concentrated and the distribution of students is fixed based on school timetable. The traditional indoor emergency evacuation models are mainly the simulation of evacuees, which focus on the microscopic behavior, lacking of description of behavior interact from the macroscopic level and route planning. Based on the research of cellular automata model and the social force model, this dissertation analyzes the interactions between people and people, people and wall, the distance to the exit. Moreover, starting from a macro perspective, the concept of "steady" is proposed to describe the evacuation, which can guide the whole evacuation. Finally, with the help of evolutionary wardrop equilibrium, the indoor equilibrium evacuation model is established by considering the individual need and system efficiency.3) Considering the characteristics of mixed traffic evacuation in the campus, this dissertation proposes a multi-objective equilibrium model for campus evacuation in road networks based on the interaction between people and vehicles. On campus, when the typical emergency events happen, people and vehicles are likely to evacuate together due to the lack of traffic lights, and easily to cause congestion, especially at traffic junctions. By fully considering the characteristics of campus road network and analyzing both the mixed flow crowding degree and the junction steering model, this dissertation establishes the emergency evacuation model of mixed traffic flow. By considering constructing the multi-objective optimal method for mixed flow, the multi-objective evacuation equilibrium model is established for mixed flow on campus.4) This dissertation develops software for indoor evacuation in large buildings, and makes experimental verifications and result analyses to the indoor evacuation model mentioned above in teaching buildings and outdoor evacuation in road network. The software is named "routes planning software for indoor evacuation in large building", referred to as CAEE, and is compared with Pathfinder. The result of the evacuation process and evacuation clearance time proves that CAEE is reasonable and effective. On this basis, CAEE is applied to Xinyi teaching building in Wuhan University of Technology to simulate the evacuation process and establish the evacuation equilibrium routes planning. After comparing with other indoor evacuation methods, the software shows the advantage in evacuation route planning. As for outdoor evacuation in the road network, the multi-objective evacuation equilibrium model is applied to the evacuation routes planning in campus of Wuhan University of Technology to carry out the simulation. Under the condition of the balanced road traffic load, statistical analyses are made on the total evacuation time and evacuation clearance time of this model. Compared with others, the results prove that the multi-objective evacuation equilibrium model is more reasonable and efficient.