设为首页
加入收藏
The Quality Control Method of Erroneous Radar Echo Data Generated by Hardware Fault
-
摘要: 雷达硬件故障直接影响数据质量，故障数据进入共享系统后不但影响本地预报员对天气系统的分析和判断，对国家级业务系统也会产生严重影响。目前，对雷达资料的数据质量控制主要针对非气象回波，对于雷达硬件故障导致的数据错误还缺乏有效的质量控制方法。该文对河北省石家庄CINRAD/SA雷达2004—2013年硬件故障时的基数据和回波特征进行分析，研究雷达故障导致的数据错误与故障类别的相关性及对数据和回波的影响。结果表明：雷达硬件故障导致的数据错误对基数据的完整性、数据位置和强度信息产生影响，发射机和接收机系统故障主要影响雷达数据的强度信息；伺服系统故障主要影响数据的位置信息。提出通过对雷达数据完整性和位置信息的检查，根据硬件故障影响雷达回波形态、位置、范围和强度等图像特征，利用基于模糊逻辑自动识别雷达硬件故障导致的错误数据的质量控制方法。利用石家庄雷达站2004—2013年雷达故障数据进行了识别效果检验，对故障数据的总体识别率超过90%，能较好实现对硬件故障导致的数据错误质量控制，是现有雷达运行正常情况下针对非气象回波的雷达数据质量控制方法的补充。Abstract: Radar hardware fault affects data quality directly. Erroneous data not only affect local forecaster analyzing weather, but also have serious influence on the running of national operation system. So far, study on radar data quality control mainly aims at non-meteorological echoes, such as ground clutter, sea clutter, electromagnetic interference and so on. There isn't enough effective quality control method for erroneous data generated by hardware fault. Through analysis on the integrity of base data, position information and characteristic of hardware fault echoes, the correlation between erroneous data and fault category, and effects of different fault on data and echoes are studied. A quality control method is provided.Erroneous data generated by radar hardware fault affect integrity of base data, position and intensity information of echoes. There is some difference among different type hardware fault or part of radar. Transmitter and receiver system fault mainly affect the intensity information. Servo system fault mainly affect position information and the integrity of date. Through checking base data integrity and echoes position information, fault data generated by servo system can be identified.The radar intensity information affect image feature such as shape, range and intensity. The error intensity information data generated by radar hardware fault can be controlled through fuzzy-logical principle, and identified through comparing parameters such as radar echoes area, mean absolute difference of intensity, the degree of intensity change, and image correlation coefficient with neighboring normal data. There is some difference between different parts of radar or different kinds of hardware fault. It is impossible to identify all erroneous data only by one method. In the actual work, it is necessary to combine status and alarm information, and apply multiple means to check radar data step by step. Only in this way, erroneous data generated by hardware fault can be effectively and comprehensively controlled.A test on erroneous data generated by hardware fault of Shijiazhuang radar site from 2004 to 2013 is carried out, and the identification ratio is above 90%. It is supplement for the existing quality control methods which mainly aim at non-meteorological echoes when radar operate normally. The proposed algorithm is mainly based on the intensity and position information, and the quality control method on velocity and spectral width error generated by hardware fault should be further studied.
-
图 1 雷达回波位置、范围、强度突变实例
Fig. 1 Cases of radar echoes position, range and intensity change
图 6 2013年4月3日04:42:00石家庄雷达回波强度
Fig. 6 Reflectivity of Shijiazhuang radar site at 044200 UTC 6 Apr 2013
图 7 2011年4月20日石家庄雷达出现饼状回波
Fig. 7 Cake shape echo of Shijiazhuang radar site on 20 Apr 2011
表 1 2013年4月3日04:42:00丝状回波位置数据
Table 1 Silk shape echo position data at 044200 UTC 6 Apr 2013
时间 方位/(°) 仰角/(°) 04:42:03.360 24.74 0.53 04:42:03.883 6.37 0.53 04:42:03.969 354.59 0.53 04:42:04.405 25.97 0.53 
下载: 导出CSV
表 2 2005年7月21—22日回波位置异常变化相关参数
Table 2 Correlation parameters of echo position change on 22 Jul 2005
时间 ΔS(≥5 dBZ) ZMAD/dBZ SPIN(≥5 dBZ) r 21T23:27—23:33 654 3.042349 0.149239 0.882581 21T23:33—23:39 1208 3.158146 0.153502 0.873645 21T23:39—22T00:45 752 3.307222 0.158321 0.858955 21T23:45—23:51 421 3.088678 0.146292 0.865209 21T23:51—23:58 918 3.061588 0.146884 0.865149 21T23:58—22T00:05 381 2.985229 0.143617 0.869761 22T00:05—00:11 704 2.828364 0.132409 0.873409 22T00:11—00:17 800 2.843678 0.132210 0.869391 22T00:17—00:23 551 2.541492 0.116812 0.886527 22T00:23—00:29 862 2.662729 0.120996 0.876078 22T00:29—00:49 543 9.305441 0.353400 0.344445 22T00:49—00:55 434 2.558339 0.117548 0.886113
下载: 导出CSV
表 3 2011年4月20日10:06—10:42相邻雷达数据的图像特征相关参数
Table 3 The correlation parameter of image features of adjacent radar data of Shijiazhuang radar site from 1006 UTC to 1042 UTC on 20 Apr 2011
时间 ΔS(≥5 dBZ) ZMAD/dBZ SPIN(≥5 dBZ) r 10:00—10:06 619 0.082349 1.315326 0.820525 10:06—10:12 238 0.081335 1.299354 0.826941 10:12—10:18 490 0.081691 1.315719 0.827493 10:18—10:24 7495 0.234046 13.341818 0.129576 10:24—10:30 7198 0.234970 13.308351 0.144123 10:30—10:36 625 0.078279 1.259607 0.841661 10:36—10:42 650 0.082204 1.326002 0.823098
下载: 导出CSV
表 4 2011年4月20日10:18—10:30回波强度变化情况
Table 4 Intensity change from 1018 UTC to 1030 UTC on 20 Apr 2011
时间 方位角/(°) 仰角/(°) 51~60 km距离库回波强度/dBZ 51 km 52 km 53 km 54 km 55 km 56 km 57 km 58 km 59 km 60 km 180.88 -2.0 1.0 -0.5 5.5 12.5 2.0 3.0 2.0 1.0 -1.5 10:18 181.85 0.53 1.5 4.5 8.5 7.5 9.0 6.0 5.5 2.5 0.0 -0.5 182.81 1.0 0.5 -0.5 6.0 8.0 11.5 8.0 -2.0 2.5 4.5 180.83 61.0 59.0 60.5 73.5 64.5 61.0 68.5 71.0 71.0 88.0 10:24 181.85 0.57 59.0 70.0 68.5 72.0 71.5 63.0 63.5 55.0 74.5 74.5 182.81 54.0 67.5 62.5 60.0 68.0 87.5 87.5 67.0 88.0 88.0 180.75 10.5 11 4.5 5.0 0.5 10.5 2.5 8.0 5.0 1.0 10:30 181.67 0.57 5.5 4 3.5 2.0 1.5 4.0 4.0 -1.5 3.5 4.0 182.64 3.5 7.5 -1.0 2.0 7.0 3.5 0.5 1.5 3.5 6.5
下载: 导出CSV
表 5 不同硬件故障数据的识别情况 (判定阈值:P≥0.6)
Table 5 Iidentification result of erroneous radar data (identification threshhold P≥0.6)
错误数据类型 回波形态 数据量 识别 识别率/% 强度异常 饼图，大范围噪点 28 22 78.57 V型缺口，扇状回波 428 394 92.06 强度异常增强或减弱 37 27 72.97 环状 1 1 100 方位角/仰角异常 回波整体方位改变 1 1 100 丝状回波 29 29 100 范围异常 2 2 100 异常数据总体识别情况 526 476 90.49
下载: 导出CSV
-
[1] 肖艳姣, 刘黎平, 杨洪平.基于天气雷达网三维拼图的混合反射率因子生成技术.气象学报, 2008, 66(3):470-473. doi: 10.11676/qxxb2008.043 [2] 肖现, 王迎春, 陈明轩, 等.基于雷达资料四维变分同化技术对北京地区一次下山突发性增强风暴热动力机制的模拟分析.气象学报, 2013, 71(5):797-816. doi: 10.11676/qxxb2013.077 [3] 刘黎平, 吴林林, 杨引明.基于模糊逻辑的分步式超折射地物回波识别方法的建立和效果分析.气象学报, 2007, 65(2):252-260. doi: 10.11676/qxxb2007.024 [4] 江源, 刘黎平, 庄薇.多普勒天气雷达地物回波特征及其识别方法改进.应用气象学报, 2009, 20(2):203-213. doi: 10.11898/1001-7313.20090210 [5] 李丰, 刘黎平, 王红艳, 等.S波段多普勒天气雷达非降水气象回波识别.应用气象学报, 2012, 23(2):147-158. doi: 10.11898/1001-7313.20120203 [6] 李丰, 刘黎平, 王红艳, 等.C波段多普勒天气雷达地物识别方法.应用气象学报, 2014, 25(2):158-167. doi: 10.11898/1001-7313.20140205 [7] 何彩芬, 黄旋旋, 丁烨毅, 等.宁波非气象雷达回波的人工智能识别及滤波.应用气象学报, 2007, 18(6):856-864. doi: 10.11898/1001-7313.200706129 [8] 谭学, 刘黎平, 范思睿.新一代天气雷达海浪回波特征分析和识别方法研究.气象学报, 2013, 71(5):962-975. doi: 10.11676/qxxb2013.074 [9] 陈忠勇, 蔡宏, 向阿勇, 等.天气雷达探测遭电子干扰分析与解决途径.气象科技, 2012, 40(5):719-722. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ201205005.htm [10] Jungbluth K, Belles J, Schumacher M.Velocity Contamination of WSR-88D and Wind Profiler Data Due to Migrating Birds//The 27th Conference on Radar Meteorology.American Meteorological Society, 1995:666-668. https://www.researchgate.net/publication/234269421_Contamination_of_Wind_Profiler_Data_by_Migrating_Birds_Characteristics_of_Corrupted_Data_and_Potential_Solutions [11] 周红根, 柴秀梅, 胡帆, 等.新一代天气雷达回波异常情况分析.气象, 2008, 34(6):112-115. doi: 10.7519/j.issn.1000-0526.2008.06.017 [12] 潘新民, 柴秀梅, 崔柄俭, 等.CINRAD/SB雷达回波强度定标调校方法.应用气象学报, 2010, 21(6):739-746. doi: 10.11898/1001-7313.20100611 [13] 胡东明, 刘强, 程元慧, 等.CINRAD/SA天线伺服系统轴角箱多次故障的分析.气象, 2007, 33(10):114-117. doi: 10.7519/j.issn.1000-0526.2007.10.017 [14] 赵瑞金, 董保华, 聂恩旺, 等.根据异常回波特征和报警信息判断雷达故障部位.气象, 2013, 39(5):645-652. doi: 10.7519/j.issn.1000-0526.2013.05.013 [15] 邵楠, 裴翀, 刘传才, 等.基于图像处理技术自动判别雷达异常产品.气象科技, 2013, 41(6):993-997. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ201306004.htm [16] 杨金红, 高玉春, 柴秀梅.新一代天气雷达运行保障能力分析.气象科技, 2014, 42(1):31-37. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ201401004.htm [17] 杨传凤, 张璇, 张骞, 等.CINRAD/SA雷达俯仰角码跳闪导致天线异动的故障排除.气象水文海洋仪器, 2011, 28(4): 109-114. http://www.cnki.com.cn/Article/CJFDTOTAL-QXSW201104033.htm [18] 蔡勤, 柴秀梅, 周红根, 等.CINRAD/SA雷达闪码故障的诊断分析.气象, 2011, 37(8):1045-1048. doi: 10.7519/j.issn.1000-0526.2011.08.019 [19] 中国气象局. 新一代天气雷达系统功能规格书 (S波段). 2010. [20] 朱永松, 国澄明.基于相关系数的相关匹配算法的研究.信号处理, 2003, 19(6):531-534. http://www.cnki.com.cn/Article/CJFDTOTAL-XXCN200306007.htm -
计量
- 摘要浏览量: 2609
- HTML全文浏览量: 1123
- PDF下载量: 760
- 被引次数: 0
