-
摘要: 为提高道路交通事故灰色预测模型的预测精度, 分析了GM (1, 1) 模型和灰色Verhulst模型的特点, 发现GM (1, 1) 模型适用于具有较强指数规律的序列, 只能描述单调的变化过程, 而Verhulst模型则适用于非单调的摆动发展序列或具有饱和状态的S形序列。针对近年来中国道路交通事故表现为具有饱和状态的S形过程, 建立交通事故Verhulst预测模型。Verhulst预测模型和GM (1, 1) 预测模型预测的2004年交通事故死亡人数分别为10.87万人和11.72万人, 相对误差分别为1.49%和9.43%, 可见Verhulst模型的预测精度明显优于GM (1, 1) 模型。
-
关键词:
- 交通安全 /
- 道路交通事故 /
- 灰色预测 /
- 灰色Verhulst模型 /
- GM(1, 1)模型
Abstract: In order to improve the predictive precisions of grey predictive models for road traffic accidents, the properties of GM (1, 1) model and grey Verhulst model were analyzed.GM (1, 1) model adapted to the sequences with good exponent law, only described monotonous change process, Verhulst model adapted swinging sequences or saturated S-shape sequences.Recently, road traffic accidents in China developed as saturated S-shape sequences, so a Verhulst predictive model of road traffic accidents was put forward.The predictive values of grey Verhulst model and GM (1, 1) model for the death people numbers of road traffic accidents in 2004 in China are (108 700) and 117 200 respectively, their predictive precisions are 1.49% and 9.43% respectively, which shows that the predictive precision of the presented model is better than that of GM (1, 1) model.-
Key words:
- traffic safety /
- road traffic accidents /
- grey prediction /
- grey Verhulst model /
- GM (1, 1) model
-
表 1 道路交通事故死亡人数统计
Table 1. Death People Numbers of Road Traffic Accidents
年份 死亡人数/万人 年份 死亡人数/万人 1990 4.93 1998 7.81 1991 5.33 1999 8.35 1992 5.87 2000 9.39 1993 6.35 2001 10.59 1994 6.63 2002 10.94 1995 7.15 2003 10.44 1996 7.37 2004 (a) 9.92 1997 7.39 2004 (b) 10.71 表 2 灰色模型精度检验等级
Table 2. Precision Grades of Grey Models
等级 指标 相对误差α 绝对关联度g0 均方差比值C0 小误差概率p0 一 0.01 0.90 0.35 0.95 二 0.05 0.80 0.50 0.80 三 0.10 0.70 0.65 0.70 四 0.20 0.60 0.80 0.60 表 3 Verhulst模型误差
Table 3. Errors of Verhulst Model
序号k 年份 原始数据xk(1)/万人 预测值万人 残差εk万人 相对误差Δk/% 1 1990 4.93 4.93 0.00 0.00 2 1991 5.33 5.35 -0.02 0.38 3 1992 5.87 5.79 0.08 1.36 4 1993 6.35 6.23 0.12 1.89 5 1994 6.63 6.69 -0.06 0.90 6 1995 7.15 7.14 0.01 0.14 7 1996 7.37 7.60 -0.23 3.12 8 1997 7.39 8.06 -0.67 9.07 9 1998 7.81 8.50 -0.69 8.83 10 1999 8.35 8.94 -0.59 7.07 11 2000 9.39 9.36 0.03 0.32 12 2001 10.59 9.77 0.82 7.74 13 2002 10.94 10.16 0.78 7.13 14 2003 10.44 10.52 -0.08 0.77 15 2004 (a) 9.92 10.87 -0.95 9.58 16 2004 (b) 10.71 10.87 -0.16 1.49 表 4 GM (1, 1) 模型误差
Table 4. Errors of GM (1, 1) Model
序号k 年份 原始数据xk(0)/万人 预测值万人 残差εk万人 相对误差Δk/% 1 1990 4.93 4.93 0.00 0.00 2 1991 5.33 5.47 -0.14 2.63 3 1992 5.87 5.80 0.07 1.19 4 1993 6.35 6.15 0.20 3.15 5 1994 6.63 6.52 0.11 1.66 6 1995 7.15 6.92 0.23 3.22 7 1996 7.37 7.34 0.03 0.41 8 1997 7.39 7.78 -0.39 5.28 9 1998 7.81 8.25 -0.44 5.63 10 1999 8.35 8.75 -0.40 4.79 11 2000 9.39 9.27 0.12 1.28 12 2001 10.59 9.83 0.76 7.18 13 2002 10.94 10.43 0.51 4.66 14 2003 10.44 11.06 -0.62 5.94 15 2004 (a) 9.92 11.72 -1.80 18.15 16 2004 (b) 10.71 11.72 -1.01 9.43 -
[1] 李相勇, 张南, 蒋葛夫. 道路交通事故灰色马尔可夫预测模型[J]. 公路交通科技, 2003, 20 (4): 98-101. https://www.cnki.com.cn/Article/CJFDTOTAL-GLJK200304026.htmLI Xiang-yong, ZHANG Nan, JI ANG Ge-fu. Grey-Markov Model for Forecasting Road Accidents[J]. Journal of Highway and Transportation Research and Development, 2003, 20 (4): 98-101. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GLJK200304026.htm [2] 周维新. 交通事故灰色预测模型的研究[J]. 西安公路交通大学学报, 2000, 20 (2): 73-75. https://www.cnki.com.cn/Article/CJFDTOTAL-XAGL200002023.htmZHOU Wei-xin. Study on the Forecast Model of Traffic Accidents with Grey Theory[J]. Journal of Xi'an Highway University, 2000, 20 (2): 73-75. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XAGL200002023.htm [3] 刘建齐, 陈兰, 刘建武. 道路交通事故预测中的灰色预测GM (1, 1) 模型[J]. 广西交通科技, 2003, 28 (4): 106-109. https://www.cnki.com.cn/Article/CJFDTOTAL-GXJK200304032.htmLIU Jian-qi, CHEN Lan, LIU Jian-wu. The Gray Forecasting GM (1, 1) Model in Forecasting Traffic Accident[J]. Guangxi Communications Science and Technology, 2003, 28 (4): 106-109. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GXJK200304032.htm [4] 韩文涛, 张倩, 贾安民. 基于灰色系统的道路交通事故预测模型研究[J]. 西安建筑科技大学学报: 自然科学版, 2005, 37 (3): 375-377. https://www.cnki.com.cn/Article/CJFDTOTAL-XAJZ200503015.htmHAN Wen-tao, ZHANG Qian, JIA An-min. Study on the Forecast Model of Road Accident with Gray Theory[J]. Journal of Xi'an University of Architecture and Technology: Natural Science Edition, 2005, 37 (3): 375-377. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XAJZ200503015.htm [5] 刘志强. 道路交通安全研究方法[J]. 中国安全科学学报, 2000, 10 (6): 14-19. https://www.cnki.com.cn/Article/CJFDTOTAL-ZAQK200006003.htmLIU Zhi-qiang. Method of Study on Road Traffic Safety[J]. China Safety Science Journal, 2000, 10 (6): 14-19. (in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-ZAQK200006003.htm [6] 孟祥海, 裴玉龙. 哈尔滨市区交通事故分析[J]. 哈尔滨建筑大学学报, 1999, 32 (4): 112-116. https://www.cnki.com.cn/Article/CJFDTOTAL-HEBJ199904024.htmMENG Xiang-hai, PEI Yu-long. Traffic Accident Analysis on Urban District of Harbin City[J]. Journal of Harbin University of Civil Engineering and Architecture, 1999, 32 (4): 112-116. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HEBJ199904024.htm [7] 盖春英, 裴玉龙. 公路货运量灰色模型—马尔可夫链预测方法研究[J]. 中国公路学报, 2003, 16 (3): 113-116. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGL200303027.htmGAI Chun-ying, PEI Yu-long. Study of the GM (1, 1) -Markov Chain Model on Highway Freight Forecast[J]. China Journal of Highway and Transport, 2003, 16 (3): 113-116. (in Chi-nese). https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGL200303027.htm [8] 周伟, 王建军, 武常吉, 等. 陕西省国民经济发展的组合预测研究[J]. 西安公路交通大学学报, 1997, 17 (3): 101-104. https://www.cnki.com.cn/Article/CJFDTOTAL-XAGL703.022.htmZHOU Wei, WANG Jian-jun, WU Chang-ji, et al. Combinative Forecast of the Economic Development in Shaanxi Province[J]. Journal of Xi'an Highway University, 1997, 17 (3): 101-104. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XAGL703.022.htm [9] 刘思峰, 郭天榜, 党耀国, 等. 灰色系统理论及其应用[M]. 北京: 科学出版社, 1999. [10] 邓聚龙. 灰色系统基本方法[M]. 武汉: 华中理工大学出版社, 1987.