摘要
针对复杂环境中多放射源定位精度低及搜索算法易陷局部极值的问题,提出基于路径规划与梯度优化的放射源定位方法。首先,利用A*算法进行全局路径规划,获取辐射强度分布数据,通过径向基函数(radial basis function,RBF)插值算法建立初步辐射场模型,采用多起点梯度上升法搜索局部极值,预测放射源位置。其次,为提高放射源定位精度与邻近源之间的区分度,建立二次优化机制,围绕初测位置进行二次路径规划密集采样,更新辐射场后实现对放射源的精确定位。实验结果证明,该方法精度高、稳定性好,可应对近距离源相互干扰及建筑遮蔽环境,为核应急响应与放射性监测提供了有效技术支撑。
To address the challenges of low localization accuracy for multiple radioactive sources in complex environments and the tendency of search algorithms to become trapped in local extrema,a radioactive source localization method based on path planning and gradient optimization was proposed.Firstly,the A*algorithm was employed for global path planning to collect radiation intensity distribution data.A preliminary radiation field model was established using the radial basis function(RBF)interpolation,followed by a multi-start gradient ascent method to search for local extrema and predict source locations.Secondly,to improve the accuracy of radioactive source localization and the discrimination between adjacent sources,a secondary optimization mechanism was implemented.Dense sampling through secondary path planning was performed around the initially identified positions,and the precise radioactive source localization was achieved after updating the radiation field.Experimental results demonstrate that the method achieves high accuracy and robust stability,and can handle the mutual interference between proximal sources and building occlusion scenarios,providing valuable technical support for nuclear emergency response and radioactive monitoring.
作者
王凯
王吉芳
马洁
秦宇飞
WANG Kai;WANG Jifang;MA Jie;QIN Yufei(College of Mechanical and Electrical Engineering,Beijing Information Science&Technology University,Beijing 100192,China)
出处
《北京信息科技大学学报(自然科学版)》
2025年第4期51-57,67,共8页
Journal of Beijing Information Science and Technology University(Science and Technology Edition)
基金
国防科技工业民用专项科研技术研究项目(KFZC2023021401)。
关键词
放射源定位
径向基函数
多起点梯度上升法
二次优化机制
radioactive source localization
radial basis function(RBF)
multi-start gradient ascent method
secondary optimization mechanism
作者简介
第一作者:王凯,男,硕士研究生;通信作者:王吉芳,女,博士,教授。