Research On Complex Scene Path Planning Based On Artificial Bee Colony Algorithm

With the world economy growing,the population increasing and the stable developing of cities,public safety problems are getting more and more.Public safety management has been caused extensive attention from all walks of life.Studying for crowd evacuation behavior and summarizing the crowd movement rule to provide the intuitive evacuation path planning,formulate a reasonable emergency plan,and ensure the safety of people's life and property.These are becoming the most important problems need to be solved.With all kinds limitation of transitional method,computer simulation technology research and application have been provided an effective solution,which overcomes the safety problems,the unreal issues and so on in the traditional way.Computer simulation technology is able to provide a realistic simulation,highly effective movement crowd behavior.3D model modeling,the crowd movement,and the macro path planning under the virtual scene are the 3 best ways to simulate the crowd behaviors.According to the above three aspects,this paper will put forward the crowd evacuation behavior model and 3D modeling method.Using 3D modeling software to build 3D scene model and virtual character model.With the disadvantages in the traditional like highly cost,difficulty operation and many other problems are getting solved.Crowd behavior simulation can effectively restore the different types of scenes and characters,and satisfy the requirements of the model modeling truthful and accurate.Due to the poor of the macro path planning algorithm applicability,crowd movement behavior problems,and do not support for complex scenarios.This paper would like to propose an improved method,which combines social force model and the artificial colony algorithm.Use the improved artificial colony algorithm for real-time macro path planning,combined with fixed social force model to guide individual movement,applied to the crowd movement simulation under different scenarios.In this paper,the main work and innovation points are as follows:To solve the topology structure of the 3d scene made by software,realize different actions of the realistic virtual characters model.Extracting the semantic information of 3d scene to obtain scene block partition relationship,thereby getting topology structure of the scene model,simplify the scene storage structure and macroscopic path planning complexity,in order to realize information acquisition and store complex scenarios.Using Motion Builder to add movement model to virtual characters,realize the virtual characters walking,running behaviors in the simulation to make the model realistic and lively,making an effective reference for the study of the crowd movement simulation.Put forward a kind of improved social force model for the motion behavior simulation of the crowd.Adding a visual impact factor to reflect the different effects of different perspectives on the individual motion,limit the forces between invisible pedestrians,in order to make a more realistic simulation of individual movement behavior.On the other hand,putting forward the temporary target strategy on obstacles bounding box vertex,implement individual obstacle avoidance behavior within visual range ahead of time,in order to actuate the pedestrian smoothly finishing local motion process.Putting forward a modified lead path selection mechanism of artificial colony algorithm based on the grouping strategy.Realize the real-time macro path planning under complex scene.Regrouping the individual based on the original group,to realize information transfer and leader selection between the same group.Using parallel computing to accelerate the algorithm convergence speed;improved leading path selection mechanism,considering both export crowded and path length.Simulate the mental phenomena of escaping.Improve the crowd real-time perception in the scene,more conform to the real crowd evacuation.Combined with undertake projects,applying the above research to the crowd motion simulation system and realistic rendering platform,in order to simulate crowd movement between series different sense.System and the platform realize the scene semantic information extraction,macro path planning,the crowd movement and realistic rendering functions.Using the system and platform to complete the simple scene and complex scene simulation experiments,to analysis and study the evacuation simulation effect.Compared with the movement behavior of the real population characteristics,we can verify the proposed method is real and effective.It has a higher reference value and application value for the future study of population behavior characteristic analysis.