摘要
针对即时战略游戏环境中传统的路径规划算法寻路因素较为单一、信息传递较少、被敌方单位击杀概率较大等问题,提出了一种仅仅基于直接交互机制的蚁群寻路算法。这种算法不仅仅是通过在蚂蚁之间直接交互信息来传递代表寻路目标点位置的信息,同时蚂蚁还与敌方单位进行直接交互,获取在路径上对己方单位带来伤害的敌方单位的位置信息,提高我方侦察单位的存活率,从而规划出一条避开敌方攻击单位到达敌方基地的更合理路径。这种算法同时还扩大了蚂蚁的感知范围,提高了蚂蚁之间进行交互的概率,缩短了完成路径规划所需的时间。通过栅格网络和游戏《星际争霸:母巢之战》仿真地图与A*算法、人工势场算法等常见的路径规划算法做对比实验,结果表明,基于直接交互机制的蚁群算法在某些栅格地图中取得了最短的路径,而且在特定游戏场景中有最长的停留时间,有效解决了游戏场景中的路径规划问题。
Traditional path planning algorithms in real-time strategy game environments often face issues such as limited factors for pathfinding,insufficient information transfer,and higher probability of enemy unit attacks.To address these challenges,this paper proposes an ant colony path planning algorithm based on a direct interaction mechanism.This algorithm utilizes direct interaction between ants to transmit information representing the location of the pathfinding target point.Additionally,ants directly interact with enemy units to obtain information about the enemy units that pose a threat to their own units along the path,thus improving the survival rate of the detection units.This approach enables the planning of more reasonable paths to avoid enemy attacking units and reach the enemy base.Furthermore,the perception range of ants is expanded,the probability of interaction between ants is increased,and the time required for path planning is reduced.Comparative experiments conducted on raster network simulation maps and the game StarCraft:Brood War demonstrates that the ant colony algorithm based on a direct interaction mechanism achieves the shortest path in certain grid maps and exhibits the longest stay time in specific game scenes,effectively addressing the problem of path planning in game scenes.
作者
孟志刚
郑佳欣
李海泉
MENG Zhigang;ZHENG Jiaxin;LI Haiquan(School of Computer Science and Engineering,Changsha University,Changsha Hunan 410022,China;School of Automotive and Mechanical Engineering,Changsha University of Science and Technology,Changsha Hunan 410004,China)
出处
《长沙大学学报》
2023年第5期1-6,共6页
Journal of Changsha University
基金
湖南省社会科学基金教育学专项课题,编号:JJ194000。
