摘要
定位精度是评价雷电定位网络的重要指标之一,定位算法直接关系到雷电探测结果的精度。经典定位法抗误差干扰能力差、定位精度低,传统迭代算法易于发散且会陷入局部最优。为了实现更有效的定位,在定位计算中引入改进密度聚类算法(adaptive density-based spatial clustering of applications with noise,ADBSCAN)。通过雷击事故实例和区域仿真分析了定位算法的性能。结果表明,ADBSCAN不需要人工干预,对于雷电定位结果的聚类效果更好;基于ADBSCAN的雷电定位算法克服了传统定位算法的缺点,提高了抗误差干扰的能力,能稳定并精确求解出雷击点。
Location accuracy is an important index for evaluation of locating networks, and the location algorithm directly related to the accuracy of the location results. Classical location method have poor ability to resist error interference and the location accuracy is low. Traditional iterative algorithms are easy to diverge and fall into local optimum. In order to achieve more effective location, an improved density clustering algorithm adaptive density-based spatial clustering of applications with noise (ADBSCAN) was introduced to location calculation. The performance of location algorithm was analyzed through lightning accident case and regional simulation. The results show that ADBSCAN does not require manual intervention, and has better clustering effect. The lightning location algorithm based on ADBSCAN overcomes the shortcomings of traditional algorithms, improves the ability to resist error interference and can solve the lightning points steadily and accurately.
作者
李涛
张帅弛
张灿
任永珍
LI Tao;ZHANG Shuai-chi;ZHANG Can;REN Yong-zhen(Electronic and Information Engineering College, Nanjing University of Information Science & Technology, Nanjing 210044,China)
出处
《科学技术与工程》
北大核心
2018年第36期53-59,共7页
Science Technology and Engineering
基金
公益性行业(气象)科研专项(GYHY201306070)
江苏省高等学校大学生创新创业训练计划(201610300031)资助
关键词
雷电定位
到达时间差
聚类分析
DBSCAN
lightning location
time difference of arrival
cluster analysis
DBSCAN
