|Author's Email Address
||This thesis had been viewed 5367 times. Download 4 times.|
||Computer Science and Engineering|
|Type of Document
||A unified approach toward crowd simulation|
|Date of Defense
||There are various kinds of creature living in the world and each kind of creatures has its own unique life habits and behavior patterns. For these reasons between the creature as well as the biology and the environment can have many interactions with each other, and we may observe these interactions easily in the frequently daily life. However the humanity may be the quantity hugest, and also have the most complex behavior at the fine race group in all kind of creatures on earth. So how could we penetrate into by observed and analysis to obtain the information which translates to the computer simulation realistically is a topic of the very hardship with challenge to presents the human behavior. |
Virtually all previous work has been agent-based, meaning that motion is computed separately for each individual. Such models can capture each person’s unique situation. So the agent-based modeling will inevitably result in a large number of calculation and make poor efficiency. In addition, there are leader-follower system which consists of a leader and multiple followers. This leader will lead the entire group to the destination, and the followers will follow the motion of their leader closely.
In our research apply the physical properties of electric charge to the simulation of pedestrians by using the basic concepts of electromagnetism. The simulation method contains the agent-based modeling and the leader-follower system at the same time. The agent-based modeling simulates pedestrian of the individual motion. The goal of the leader-follower system is to simulate the real world behavior that people in a community often move by following a specific object.
||Chung-Nan Lee - chair|
Jung-Jae Chao - co-chair
John Y. Chiang - advisor
indicate in-campus access in a year and off_campus not accessible|
|Date of Submission