摘要
雷暴是威胁电力系统安全可靠运行的重要因素,监测雷暴活动可以为电力系统雷电防护提供雷暴活动路径及预警信息。该文提出一种组合八邻域雷暴跟踪算法及跟踪结果评估指标,并采用该算法对珠三角地区九次雷暴活动进行时空演变特征分析和定量评估。结果表明,九个雷暴活动共识别出大于18 min的轨迹1490条,其中简单轨迹和含分裂合并过程的复杂轨迹各占一半。雷暴活动平均速度为51.4 km/h,速度中值为43.0 km/h,简单轨迹平均速度比复杂轨迹慢。评估指标命中率(POD)、虚假警报率(FAR)和临界成功指数(CSI)分别为64.3%、40.4%和42.7%,轨迹时长中值、平均轨迹偏差、非连贯性和平均合并分裂次数分别为24 min、2.80 km、0.0327 s^(-1)和2.51次。该方法可以很好地描述雷暴活动时空演变过程和评估跟踪结果,为雷电预警提供了有效手段。
Lightning is the main cause of damages to power systems.The planning of power system transmission corridors and line lightning protection are mainly based on the lightning density and the distribution of lightning days.Measures such as reducing the grounding resistance of towers and installing lightning arresters are taken to improve the insulation level of transmission lines.The traditional lightning density and the number of thunderstorm days are not able to present the temporal and spatial evolution characteristics of thunderstorm activities.Monitoring the movement of lightning activity can provide track and warning information for lightning protection of equipment.This paper presents a combinatorial eight-connectivity tracking method and evaluation indexes based on total lightning data,which is applied to analyze the temporal and spatial evolution of lightning activities in Guangzhou.The tracking method based on total lightning data is divided into two steps:cluster identification and cluster tracking.At the step of identification,the grid method is used to pre-process the total lightning events and the lightning region(112°E~115°E,22°N~25°N)is gridded into 0.01°×0.01°grid boxes(longitude×latitude).Total lightning events are distributed into grid cells according to their geographical coordinates to form a total lightning density map in a single time interval.First,the electrically active region is identified by the 8-connectivity method and form primary lightning clusters.According to the size of the transmission line span,primary lightning clusters are merged into final lightning clusters within the distance of two grid cells.At the step of tracking,the algorithm takes influencing factors into account to match clusters between two successive time intervals,including history velocity,lightning cluster area and split-merge processes.In addition,the evaluation indexes contain the skill scores of the probability of detection(POD),false alarm ratio(FAR),and critical success index(CSI)from the contingency table approach,the track characteristics of track duration,track linearity,track continuity and track complexity to quantitatively evaluate the performance of tracking method.The tracking method and evaluation are used to analyze the nine thunderstorms and case 1 are presented as an example.The case 1 has 313 tracks longer than 18 min,and tracks move in the northeast direction,with a mean speed of 51.4 km/h and a median speed of 43.0 km/h.The centroid of lightning clusters and temporal and spatial evolution characteristics of each track are counted.The evaluation indexes of POD,FAR and CSI are 63.3%,41.4%and 43.8%,respectively,while the median track duration,mean jitter of track,incoherence and average number of splits and merges are 24 min,2.85 km,0.0341 s-1 and 2.43,respectively.The following conclusions can be drawn from the nine thunderstorms based on the combinatorial eight-connectivity tracking method and evaluation indexes:(1)A total of 10,894 lightning clusters were identified in the nine lightning activities,with a median area of 22 km2.It is found that 1490 lightning cluster tracks are longer than 18 min,of which simple tracks and complex tracks(involving split-ting or merging)account for half,respectively.(2)Six lightning activities move in the northeast direction,the rest move northwest,with a mean speed of 51.4 km/h and a median speed of 43.0 km/h.Simple tracks move slower than complex tracks.(3)POD,FAR and CSI are 65.6%,39.2%,46.4%,respectively,while the median track duration,mean jitter of track,incoherence and average number of splits and merges are 24 min,2.80 km,0.0327 s^(-1) and 2.51,respectively.This algorithm can describe the temporal and spatial evolution of lightning activity and indexes are able to evaluate the tracking results,which provides an effective method for predicting thunderstorm movement and lightning warning of transmission lines.
作者
黄怡鋆
樊亚东
王红斌
蔡力
王建国
Huang Yijun;Fan Yadong;Wang Hongbin;Cai Li;Wang Jianguo(Engineering Research Center of Lightning Protection&Grounding Technology Ministry of Education,Wuhan 430072 China;School of Electrical Engineering and Automation Wuhan University,Wuhan 430072 China;Guangzhou Electric Power Supply Bureau,Guangzhou 510013 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2024年第5期1536-1547,共12页
Transactions of China Electrotechnical Society
基金
国家自然科学基金(52177154)
中央高校基本科研业务费专项资金(2042022kf0016)资助项目。
关键词
雷暴活动
跟踪算法
评估指标
时空演变
组合八邻域
Lightning activity
tracking method
evaluation index
spatial-temporal evolution
combinatorial eight-connectivity
