Research On Path Searching Based On Terrain Analysis

It is the central motive and key method to investigate the behave of mankind in the simulation of virtual battlefield. Computer generated forces are the computer actors in the virtual environment. They imitate the behave of mankind making proper respond to the events in the virtual battlefield, so computer generated forces are needed to have certain brain. Independence of motion is the specific feature of mankind and path planning is the basic intelligence of mankind, so it is necessary to carry out the research of path planning for computer generated forces. At present, in the virtual world and robot domain related with artificial intelligence, there are some perfectly effective algorithms for the path planning. With the improvement of requirement for path planning, some excellence algorithms such as A* algorithm and Genetic Algorithm have been widely used in the solution of path planning problem, but the achievements are single aim optimization pathes leaving out the effect of terrain. In this paper, on the basis of analysis and comparison of these two algorithms, according to the practical problems the author expounds the A* algorithm based on terrain effect, and accomplishes multi-aim optimization path planning. In the specific application, the author accomplishes three procedure of path planning: first, analysis the terrain environment and remove the information of elevation and vegetation. Secondly, selects new heuristic function of A* algorithm and introduces the terrain information into the algorithm, in the test, visualizes the result using graphic develop kit and compare it with the result acquired the traditional A* algorithm, at the same time, compares the results of different proportion assignment, achieving the abundance of path planning. Finally, integrates the algorithm as a part of system into the virtual battlefield system, commanding computer generated forces. The movements of computer generated forces are showed through the 2D and 3D man and machine interface. The algorithm is decisive for the research of virtual battlefield, and it is a reference for the path planning in the virtual environment.