摘要
为了研究传统蚁群算法在水下智能机器人(AUV)的三维路径规划中由于算法本身的原因,初始信息素信息匮乏,路径搜索规划速度慢,容易陷入某个局部最优的状态等问题,提出一种改进型蚁群—粒子群融合算法,充分利用粒子群算法较强的全局搜索能力。在环境模型的基础上建立干扰模型,将障碍与探测作为突破点,提出代价函数用于蚁群算法的优化中。设计了新的信息素分布(包括更新和挥发)和启发函数,在转移概率公式中加入区域安全因素,提高了三维路径规划的准确度;为了提高三维路径规划的速度,在传统粒子群算法基础上优化权重函数。最后进行仿真,结果证明该方法有效可行。
The reason why the inadequate initial information, and slow to search path planning, easy to fall into a local optimal state, is that the optimization itself that the traditional ant colony in intelligent underwater robot（AUV）. It was proposed an improved ant colony optimization and particle swarm optimization. That was made full use of the strong ability of ant colony optimization in the whole process. Not only on the basic of the environment model that was established the interference model, and was brook through the barrier and detection, but also the cost function was put forward for the optimization of ant colony optimization. The design of a new pheromone distribution（including updating and evaporation）and heuristic function. And we was added regional security factors in the transition probability formula that improved the accuracy of three-dimensional path planning. In order to improve the speed of three-dimensional path planning that was optimization of weight function in the traditional particle swarm algorithm. Simulation experiments prove the feasibility and validity of this method.
作者
付振秋
季光
杨瑛
FU Zhen-qiu;JI Guang;YANG Ying(AVIC China Aero Polytechnology Estabilishment,Beijing 100028,China)
出处
《舰船科学技术》
北大核心
2018年第10期72-77,共6页
Ship Science and Technology
关键词
三维路径规划
AUV
蚁群算法
粒子群算法
干扰模型
three-dimensional path planning
AUV
ant colony optimization
particle swarm optimization
inter-ference model
