One Car Hybrid Environment Research And Implementation Of Intelligent Control

Group animation has been an important research field of computer direction, at the same time and artificial intelligence are closely linked, with the extensive application of the video, animation, games and many other fields. Through group animation to simulate different scenarios, group behavior, through the lesser cost to obtain a variety of available research data, can reproduce the scene, or anticipatory demo, without the use of a lot of manpower and material resources, disaster simulation, city planning has a great help. Group animation in a more important research is the intelligent routing, how to express different scenarios of mass movement situation, make it more close to the real motion, is a research focus of group animation.Common groups of simulation intelligent control are mostly focused on the micro agent based approach, or macro based on continuous mode, this thesis on the intelligent control method to do research. Agent based approach has its advantages. In general, while each individual units in a group environment, but their awareness is relatively independent, they have their own judgment, although by groups of environment influence, but his personality is also need to be recognized. Although the agent based approach has its advantages, but in the mass simulation conditions, cost is still very big. Based on the study of intelligent control of continuous way, the predecessors have done a lot of work, proposed a variety of methods. Based on the dynamic potential field based real-time group model, integrated into the moving obstacle (such as a vehicle) of the global navigation, effectively solve the problems of mass motion control, but does not require the calculation of collision avoidance.Based on the existing project to study, for a particular environment simulation requirement, seek to people and vehicles strong interactive simulation method. Found in continuous mode based on simulation method, dynamic potential field method and gas dynamics method in speed, density is calculated to have similar place, but are macroscopic control method. Dynamic potential field is more suitable for the large-scale population movement control; gas dynamics method is more suitable for the single track vehicle motion control.This thesis focuses on the dynamic potential field method and gas dynamics method to do the further study, around the crowd scene vehicle hybrid intelligent control subjects, and in combination with the original project in the simulation model, the following work:Firstly, the dynamic potential field routing method, according to the original project simulation model did adaptation, which can be applied to the original group animation simulation system, and the simulation system for collision avoidance of combined modalities, achieve the macro control and micro control purposes, at the same time, the potential field optimization size method used in vehicle control, replace the original vehicle cruise system fixed point method, so that the vehicle motion control more flexible. Secondly, the dynamic potential field in the velocity control strategy is introduced into the original system, so the system of the crowd to a certain extent in the autonomic regulation of speed, so that it is more close to the truth. Secondly, the aerodynamic method, is applied to the original simulation system for vehicle speed control, the vehicle is more smooth and real. Finally, through the construction of congestion path method, the aerodynamics of speed control method is applied to the congestion control; potential field path leads to congestion problems, while avoiding the global potential refresh too high a price problem. Through the above method, the simulation system for people and vehicle path and velocity control more flexible and real, can be adapted to more of the scene, at the same time, can make people and vehicles of more interactive.