欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2017, Vol. 38 ›› Issue (3): 520-526.doi: 10.3969/j.issn.1000-1093.2017.03.014

• 论文 • 上一篇    下一篇

基于凸优化的水下载体定位研究

曹俊1,2, 郑翠娥1,2, 孙大军1,2, 张殿伦1,2   

  1. (1.哈尔滨工程大学 水声工程学院, 黑龙江 哈尔滨 150001; 2.哈尔滨工程大学 水声技术重点实验室, 黑龙江 哈尔滨 150001)
  • 收稿日期:2016-07-04 修回日期:2016-07-04 上线日期:2017-04-24
  • 通讯作者: 张殿伦(1967—), 男, 教授, 博士生导师 E-mail:zhangdianlun@hrbeu.edu.cn
  • 作者简介:曹俊(1987—), 男, 博士研究生。E-mail: caojun1105@sina.cn
  • 基金资助:
    国家自然科学基金项目(61531012);国家国防科技工业局技术基础科研项目(B2420133002)

Research on the Positioning of Underwater Vehicle Based on Convex Optimization

CAO Jun1,2, ZHENG Cui-e1,2, SUN Da-jun1,2, ZHANG Dian-lun1,2   

  1. (1.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, Heilongjiang, China;2.Acoustic Science and Technology Laboratory, Harbin Engineering University, Harbin 150001, Heilongjiang, China)
  • Received:2016-07-04 Revised:2016-07-04 Online:2017-04-24

摘要: 基于单信标测距的定位方法,是水声定位技术一个新的研究方向,具有重要应用价值。针对单信标测距定位问题最小二乘结构,对高度非凸的测距定位方程首次提出了基于凸优化的解算方法。对带有误差的测距量测方程添加不等式约束条件,将其改写为有约束的解算方程;对目标函数和约束条件进行放射近似,将其进行凸化处理;引入松弛变量和罚函数,降低对迭代初值的敏感度。仿真结果表明:对于不同量级的初值误差,定位结果的差别在厘米量级,所提算法不需要可靠的初值即能迭代收敛得到高精度的定位结果;信标与圆形航迹的圆心距离越近,定位精度越高。通过湖上试验,进一步验证了所提算法的有效性。

关键词: 信息处理技术, 单信标测距, 虚拟信标, 凸优化, 约束条件, 罚函数

Abstract: The localization method based on single beacon ranging is a new research direction of underwater acoustic positioning technology. A method based on convex optimization is proposed for the first time to solve the least square structure for the single beacon ranging and locating. An inequality constraint condition is added to the ranging equation with errors, and the ranging equation is reformulated as a constrained solution equation. The objective function and constraint condition are approximated by radiation, and the convex processing is carried out. The slack variable and penalty function are introduced to reduce the sensitivity of iteration initial value. Simulated results show that the difference of positioning results is in centimeter order of magnitude for different magnitudes of initial errors. The proposed algorithm can be iterated convergence to get high accuracy of the positioning results without reliable initial value. When the beacon is close the center distance of circular track, the positioning accuracy is higher. The effectiveness of the proposed algorithm is verified by experiments on the lake.Key

Key words: informationprocessingtechnology, singlebeaconranging, virtualbeacon, convexoptimization, constraintcondition, penaltyfunction

中图分类号: