Simulation Of Crowd Behavior Based On Genetic Algorithm

With the development of virtual reality technology, especially the rapid development of the digitized human behavior simulation technique, in virtual environment the simulation technique of large-scale crowd behavior started to become a virtual reality technology important research direction, simultaneously it is also an important support technology to enhances the virtual environment reality and the immersion, and it has quite wide ranges of application value in education, entertainment, public security domain and large areas of design and so on. At the present stage of the study, the group behavior simulation's main purpose displays a overall behavior features of large groups in a scene, what pursues is the visible reality, the main object is the specific place large-scale gregariousness groups behavior research, lacks the versatility.The research in simulation of crowd behavior related to computer graphics, simulation technology, multimedia technology, artifical intelligence and other related technology, it is actually more than one goals.constrained combination programming problem, how to use limited resources reasonable arrangement behaviors for crowd each individual, time, place, so that the overall simulation of crowd behavior is orderly and realistic. But the genetic algorithm is highly parallel, randomized, adaptive search capabilities strong features, particularly suited to dealing with traditional search algorithms not solved problems of complex and nonlinear. This article attempts to apply genetic algorithm to solve the problem of crowd behavior simulation, and designed a special kind of objective evaluation function and genetic encoding scheme. Experiments show that the method can greatly reduce the search space, and enables optimal results. This study is mainly from the following aspects:1. For the crowd simulation based on genetic algorithm research background and significance are introduced, while analysis of group behavior at home and abroad development of simulation technology, introduced the crowd behavior simulation and genetic algorithms relevant basic theoretical knowledge.2. For the combinatorial optimization problem of crowd behavior simulation multi-objective and constrained, proposed a genetic algorithm based on crowd behavior simulation program. By setting the special data structure to simulate the process of a variety of constraint rules, genetic encoding of genetic algorithm, fitness evaluation function implement of crowd behavior simulation. Then experiment verified the rationality and-feasibility of the algorithm.3. A crowd behavior simulation system is established, which is set according to user requirements of different parameters to achieve the right of crowd behavior simulation. At the same time, using an object-oriented idea to make the system has a very good versatility and expandability.