Research On CGF Tactical Path Planning In Combat Element Simulation

With the widespread application of Computer Generated Forces(CGF) in the military simulation, behavior modeling of CGF agent has challenged CGF researchers and developers for years. Among the various CGF behaviors, path planning plays a basic role to perform "real behavior" because it’s the executing precondition of many high level intelligent behaviors. However, lacking consideration for combat environment and tactical requirement, traditional path planning conducts an unreal behavior performance. Thus it needs to pay more attention on the influence of combat element environment and tactical requirements to improve the intelligence and authenticity of path planning.To research the CGF path planning problem in combat element simulation, firstly we need to describe the problem accurately. Based on the structure of agent, a tactical path planning frame is proposed for CGF agent. According to the frame, the natural environment and threat modeling is conducted for combat element environment. In which, the potential field method and probabilistic threat exposure map method are adopted to build the threat situation model. Finally we carry out the path optimization formula of length and danger degree, and express the CGF tactical path planning problem as a normal optimization problem.According to the degree differences of environment information agent obtains, local and global path search methods are researched respectively. In the method of local tactical path planning based on the artificial potential field(APF), angle conditions and memorial wall-following path are introduced as the new condition that agent gets away from the wall-following mode to solve the local minimum in traditional APF method, adapting the complex environment with non-convex polygon obstacles. Besides, to make the APF method suitable for the combat element tactical path planning, we propose an improved APF method integrating the traditional obstacle potential field with the threat potential field, adjust it with environment, situation and tactical requirements. In the method of global tactical path planning based on particle swarm optimization(PSO), an adaptive polar coordinate with unequal length is designed to make the search space cover the useful waypoints behind the obstacles most. Learning from idea of gravitational search algorithm(GSA), we add the fitness information of particle into the velocity update formula, and adopt a performance compared method considering constrained dominance relationships of particles, improving the speed and precision of search algorithm. At last, the integrated path evaluation considering the trafficability and masking of the terrain and obstacle is used to guide path search, making planned path meet the requirements of environment, situation and tactical rule.Simulation results demonstrate the validity of the two algorithms proposed. The improved artificial potential field method can solve the local minimum problem in complex obstacle environment well, and the planned path meets the requirements of avoiding threat and making full use of obstacles as blindage. The CGF tactical path planning method based on the particle swarm optimization reflects accurately the influence of terrain and situation on the path choose, plans a safer and shorter path.