摘要
大气相干长度是度量大气湍流对光波前影响程度的重要参数,在多个领域中具有关键意义.为提高其测量精度,提出一种基于六孔径K-means聚类算法的大气湍流特征参数提取方法,解决了在测量过程中出现光斑边界不清晰甚至粘连的问题.此外,建立了大气相干长度测量系统,系统选用小口径光学天线和六孔径掩膜装置,在增加了统计值的同时兼顾了系统的便携性.利用该系统对长春地区进行多日全天实验监测,其实验数据表明,全天测量呈现出的趋势及数值符合湍流强度变化规律,所提方法相较于传统的阈值分割算法的精确性提升了 33.05％,相关实验可为长春地区的湍流监测提供技术手段.
Abstract
Objective Due to its advantages of low cost,high speed,and strong anti-interference capabilities,laser transmission technology has promising applications in both civilian and military fields.However,laser links inevitably pass through atmospheric channels during transmission.The scintillation and absorption effects in the atmosphere can cause laser signal attenuation and angle-of-arrival fluctuations,ultimately affecting the quality of the communication link.In severe cases,this may lead to link interruption and communication failure.Therefore,effectively,accurately,and in real-time measuring the dynamic characteristics of atmospheric coherence length is of great importance for the practical engineering applications of laser transmission.Methods In this study,we developed an atmospheric coherence length measurement system based on a six-aperture DIMM(Differential Image Motion Monitor).Both ends of the system use small-aperture optical antennas.At the emitting end,the laser is expanded and the system's focal length is adjusted to ensure the laser exits as parallel light.At the receiving end,a six-aperture mask device is placed in front of the optical antenna.After passing through a 532 nm narrowband filter,the light is focused and imaged onto the focal plane of a CMOS camera within the optical antenna.The spots in the images are then processed using a K-means clustering boundary identification algorithm.The results are calculated on a PC and finally displayed on the upper computer interface.Results and Discussions This study proposes a parameter extraction method for a six-aperture atmospheric turbulence characteristic measurement system based on the K-means clustering algorithm(Fig.2).This method can address issues of spot adhesion and boundary blurring.Compared to the traditional threshold segmentation algorithm,the identification accuracy has been improved from 66.873％to 99.923％(Fig.5).Finally,a six-aperture differential image motion experiment(Fig.6-Fig.7)was conducted to verify the system's feasibility and the algorithm's stability.The atmospheric coherence length values measured using the threshold segmentation algorithm exhibit significant calculation errors under moderate to strong turbulence,demonstrating the stability of the proposed algorithm(Fig.9).Additionally,the turbulence intensity in the Changchun area was measured.The experimental data showed that the trends and values observed throughout the day were consistent with the known patterns of turbulence intensity variation(Fig.10),further confirming the effectiveness and feasibility of the system and method used in this study.Conclusions This study constructed an atmospheric coherence length measurement system based on a six-aperture design,which enhances statistical reliability while maintaining system portability.To address issues such as unclear and overlapping speckle boundaries during measurement,a six-aperture K-means clustering algorithm was developed for extracting atmospheric turbulence characteristic parameters.Compared to the traditional threshold segmentation method,this new approach increases recognition accuracy from 66.873％to 99.923％.Through continuous day-long observations over several days,the resulting atmospheric coherence length variation curves align with the expected patterns of turbulence intensity changes.The system and method presented in this study enable real-time,continuous measurement of atmospheric coherence length.The designed upper computer interface allows for real-time display and ease of operation,making it applicable in practical engineering applications.
维普
万方数据