列车轨道占用自动识别算法

蔡伯根; 严细辉; 王剑; 上官伟

doi:10.19818/j.cnki.1671-1637.2010.06.018

列车轨道占用自动识别算法

doi: 10.19818/j.cnki.1671-1637.2010.06.018

蔡伯根^1,,
严细辉¹,
王剑^{1, 2},
上官伟^{1, 2}

1.
北京交通大学电子信息工程学院, 北京 100044
2.
北京交通大学轨道交通控制与安全国家重点实验室, 北京 100044

基金项目:

国家自然科学基金项目 60736047

国家自然科学基金项目 60870016

国家863计划项目 2009AA11Z221

轨道交通控制与安全国家重点实验室自主研究课题 RCS2009ZT013

中央高校基本科研业务费专项资金项目 2009JBM005

铁道部科技研究开发计划项目 2009X003-G

详细信息

作者简介:
蔡伯根(1966-), 男, 江苏如皋人, 北京交通大学教授, 工学博士, 从事交通信息工程及控制研究

中图分类号: U284.48
计量
- 文章访问数: 780
- HTML全文浏览量: 106
- PDF下载量: 651
- 被引次数: 0
出版历程
- 收稿日期: 2010-06-27
- 刊出日期: 2010-12-25

Automatic identification algorithm of train track occupancy

CAI Bo-gen^1
,,
YAN Xi-hui¹,
WANG Jian^{1, 2},
SHANGGUAN Wei^{1, 2}

1.
School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China
2.
State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing 100044, China

More Information

Author Bio:
CAI Ba-igen(1966-),male, professor,PhD,+86-10-51687111,bgcai@bjtu.edu.cn

摘要

摘要: 为解决列车在道岔及平行股道区段的轨道占用自动识别问题, 基于LTS-Hausdorff距离, 结合D-S证据理论, 提出了一种新的列车轨道占用自动识别算法, 建立了可用于列车轨道占用自动识别的轨道LTS-Hausdorff距离参考模板, 分析了LTS-Hausdorff距离的计算过程及轨道占用自动识别决策方法, 研究了列车速度与搜索阈值对自动识别算法的影响。验证结果表明: 在轨迹点数量为10时, 该识别算法和基于最大似然准则的轨道识别决策的识别结果相同; 列车速度越高, 轨迹点越少, 算法仍可进行有效识别; 搜索阈值越小, 算法实现时间越短。可见, 识别算法有效。
- 交通控制 /
- 轨道占用识别 /
- LTS-Hausdorff距离 /
- D-S证据理论 /
- 列车定位
Abstract: In order to resolve the automatic identification problems of train track occupancy at turnouts and on parallel sections, a new automatic identification algorithm was proposed based on LTS-Hausdorff distance and D-S evidence theory. The reference template of track LTS-Hausdorff distance was established, the calculation process of LTS-Hausdorff distance and the decision method of automatic identification were analyzed, and the effects of train speed and search threshold on the algorithm were studied. Test result shows that when there are 10 track points, the results of the new algorithm and the maximum likelihood track identification decision are same. The higher train speed is, the less track points are, and the algorithm is still effective. The smaller search threshold is, the shorter the algorithm realizing time is. So the algorithm is valid.
- traffic control /
- track occupancy identification /
- LTS-Hausdorff distance /
- D-S evidence theory /
- train positioning

HTML全文

图 1 轨道占用自动识别过程

Figure 1. Automatic identification process of track occupancy

下载: 全尺寸图片幻灯片

图 2 列车轨迹采样点

Figure 2. Sampling points of train trajectory

下载: 全尺寸图片幻灯片

图 3 T_max对搜索时间的影响

Figure 3. Influence of T_max on searching time

下载: 全尺寸图片幻灯片

图 4 T_max对LTS-Hausdorff距离计算的影响

Figure 4. Influence of T_max on LTS-Hausdorff distance calculation

下载: 全尺寸图片幻灯片

图 5 正线样本累加概率

Figure 5. Cumulative probability of main track sample

下载: 全尺寸图片幻灯片

表 1 道岔区段轨道识别

Table 1. Track identification on switch sections

样本类型	参考模板Ⅰ	参考模板Ⅱ	D-S识别结果
正线	-16.831	-100.410	正线
	-17.952	-98.451
	-16.735	-75.105
侧线	-89.364	-19.352	侧线
侧线	-96.350	-20.382	侧线

下载: 导出CSV

表 2 平行股道区段轨道识别

Table 2. Track identification on parallel track sections

样本类型	模板Ⅰ	模板Ⅱ	模板Ⅲ	D-S识别结果
第7股道样本	-20.63	-150.40	-512.80	第7股道
第8股道样本	-153.00	-15.36	-149.60	第8股道
第9股道样本	-510.90	-151.10	-16.64	第9股道

下载: 导出CSV

表 3 不同列车速度对轨道识别的影响

Table 3. Influences of different train speeds on track identification

车速/ (m·s^-1)	5 (正线样本)	10 (侧线样本)	25 (正线样本)
正线模板	-159.665	-143.770	-7.537
侧线模板	-500.870	-127.230	-11.548

下载: 导出CSV

表 4 不同列车速度下的融合结果

Table 4. Fusion results at different train speeds

车速/ (m·s^-1)	5 (正线样本)	10 (侧线样本)	25 (正线样本)
正线模板	0.651	0.235	0.673
侧线模板	0.183	0.712	0.351

下载: 导出CSV

参考文献(14)

[1]	王剑, 张辉, 蔡伯根, 等. 基于GNSS的轨道占用识别技术综述[J]. 铁道学报, 2010, 32 (2): 93-97. https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB201002018.htm WANG Jian, ZHANG Hui, CAI Bai-gen, et al. ASurvey oftrack occupancy recognition technology based on GNSS[J]. Journal of the China Rail way Society, 2010, 32 (2): 93-97. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB201002018.htm
[2]	SAAB S S. A map matching approach for train positioningpart I: development and analysis[J]. IEEE Transactions on Vehicular Technology, 2000, 49 (2): 467-475. doi: 10.1109/25.832978
[3]	王剑, 张辉, 蔡伯根, 等. 基于HMM的列车轨道占用识别中算法研究[J]. 铁道学报, 2009, 31 (3): 54-58. https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB200903011.htm WANG Jian, ZHANG Hui, CAI Bai-gen, et al. The algo-rithm of automatic track occupying identification based onHMM[J]. Journal of the China Rail way Society, 2009, 31 (3): 54-58. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-TDXB200903011.htm
[4]	谭青, 向阳辉. 加权证据理论信息融合方法在故障诊断中的应用[J]. 振动与冲击, 2008, 27 (4): 112-116. https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ200804030.htm TAN Qing, XIANG Yang-hui. Application of weighted evidential theory and its information fusion method in fault diagnosis [J]. Journal of Vibration and Shock, 2008, 27(4): 112-116. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZDCJ200804030.htm
[5]	BUSTOS J P, DONOSO F, GUESALAGA A, et al. Matchingradar and satellite i mages for ship trajectory esti mation usingthe Hausdorff distance[J]. IET Radar, Sonar and Navigation, 2007, 1 (1): 50-58. doi: 10.1049/iet-rsn:20060025
[6]	AHN Y J. Hausdorff distance between the offset curve ofquadratic bezier curve and its quadratic approxi mation[J]. Communications of the Korean Mathematical Society, 2007, 22 (4): 641-648.
[7]	LU Yue, TAN C L, HUANG Wei-hua, et al. An approach to word image matching based on weighted Hausdorff distance[C]//IEEE. Sixth International Conference on Document Analysisand Recognition. Seattle: IEEE, 2001: 921-925.
[8]	KANGJian-xin, QI Nai-ming, HOUJian. A hybrid method combining hausdorff distance, genetic algorithm and simulated annealing algorithm for image matching[C]//IEEE. ICCMS 10 Proceedings of the 2010 Second International Conference on Computer Modeling and Si mulation. Sanya: IEEE, 2010: 435-439.
[9]	SHANG Fu-hua, WANG Hui, MA Nan. A method of depth correction of logging curves based on LTS Hausdorff distance[C]// IEEE. 2010International Conference on Computer Application and System Modeling. Taiyuan: IEEE, 2010: 108-112.
[10]	SHEN Shu-han, SHI Wen-huan, LI U Yun-cai. Monocular3-D tracking of inextensible deformable surfaces under L₂-Norm[J]. IEEE Transactions on I mage Processing, 2010, 19 (2): 512-521.
[11]	SCHMITT L, MEYK H. Asystematic framework for iterativemaxi mumlikelihood receiver design[J]. IEEE Transactionson Communications, 2010, 58 (7): 2035-2045.
[12]	谷正气, 胡林, 黄晶, 等. 基于改进D-S证据理论的车辆导航地图匹配[J]. 汽车工程, 2008, 30 (2): 141-145. https://www.cnki.com.cn/Article/CJFDTOTAL-QCGC200802010.htm GU Zheng-qi, HULin, HUANGJing, et al. Vehicle naviga-tion map matching based on modified D-S evidence rule[J]. Automotive Engineering, 2008, 30 (2): 141-145. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-QCGC200802010.htm
[13]	SUDANO J J. Inverse pignistic probability transforms[C]∥IEEE. Proceedings of the Fifth International Conference on Information Fusion. Indianapolis: IEEE, 2002: 763-768.
[14]	王剑. 基于GNSS的列车定位方法研究[D]. 北京: 北京交通大学, 2007. WANG Jian. Research of the train positioning technologies based on GNSS technology[D]. Beijing: Beijing Jiaotong University, 2007. (in Chinese)