基于混合粒子群优化的海底地形辅助导航算法  

Seabed Terrain Aided Navigation Algorithm Based on Hybrid Particle Swarm Optimization

作　　者：袁赣南[1] 谭佳琳 尹伟伟 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《计算机仿真》2010年第8期122-126,共5页Computer Simulation

基　　金：国家自然科学基金项目(60704001)

摘　　要：由于海底地形的独特特性,将传统地形匹配算法直接应用于海底地形匹配时,常常会出现定位精度大幅降低的现象,而且传统遍历搜索的庞大计算量也大大降低了地形匹配的效率。为提高匹配效率,提出了一种新的算法,关于混合粒子群优化的海底地形匹配算法。算法充分利用Hausdorff距离强抗干扰能力和容错能力的特点,采用平均Hausdorff距离作为相似性测度,并在搜索策略上,用了粒子群算法来加快搜索的速度,并提出将混沌局部搜索嵌入到粒子群算法中,形成了混沌思想的混合粒子群优化算法。通过某海域电子海图的海底地形匹配实验结果证明,算法能够快速求得最优解,使系统匹配误差(CEP)较传统方式降低了13%,计算量减少了18%,是一种很有开发和研究价值的海底地形匹配算法。When most of terrain matching algorithms were directly applied to seabed terrain aided navigation system（STAN）,the positioning accuracy fell off sharply and the system became unstable because of the particularity of seabed terrain,and the efficiency was low for large amount of calculation.A new approach of seabed terrain match algorithm was presented.It used the average Hausdorff distance as similarity measure for its great anti-interference and fault-tolerance performance.As to searching strategy,hybrid particle swarm optimization algorithm was used to achieve high computing speed.The experiments based on electronic chart confirmed the availability of the proposed approach.The algorithm was robust,and it can reduce the computational complexity by 18%.The positioning accuracy was increased by 13%.

关 键 词：地形匹配 粒子群优化 地形辅助导航 水下导航 

分 类 号：U666.11[交通运输工程—船舶及航道工程]