﻿ 海上船舶交通流行为特征提取算法
 舰船科学技术  2022, Vol. 44 Issue (19): 42-45    DOI: 10.3404/j.issn.1672-7649.2022.19.009 PDF

1. 浙江国际海运职业技术学院 航海工程学院，浙江 舟山 316021;
2. 江苏航运职业技术学院，江苏 南通 226010

Design of feature extraction algorithm for marine vessel traffic
LI Zi-qiang1, ZHAO Cang-long2
1. Zhejiang International Maritime College, School of Navigation Engineering, Zhoushan 316021, China;
2. Jiangsu Shipping College Marine Technical Institute, Nantong 226010, China
Abstract: Waterway transport plays an important role in China's commodity trade and logistics. With the development of economy, the number of inland shipping and maritime transport ships in China is constantly reaching new heights. In order to strengthen the supervision level of inland shipping and maritime shipping traffic management departments for ships, and improve the traffic safety of ships in the waterway, it is of great significance to study the characteristics of ship traffic prevalence. This paper introduces the composition of the feature extraction system of marine vessel traffic epidemic from the aspects of the basic features and extraction algorithm of marine vessel traffic epidemic, and expounds the algorithm principle of the feature extraction of marine vessel traffic epidemic.
Key words: traffic flow     behavior characteristics     shipping management     algorithm
0 引　言

1)改善水上航运交通

2)助力海上交通学科理论发展

3)提高海上航运交通的安全性[1]

4)指导港口和航道规划建设

1 海上船舶交通流行为特征的概念模型及拟合算法 1.1 海上船舶交通流行为特征的概念模型

 图 1 海上船舶交通流行为特征示意图 Fig. 1 Schematic diagram of prevalence of marine ship traffic

1)交通流速度

 ${V_0} = \dfrac{l}{{\displaystyle\sum\limits_{i = 1}^n {\dfrac{{{t_i}}}{n}} }} \text{。}$

2)流量

 ${Q_0} = \frac{1}{n}\sum\limits_{i = 1}^n {{q_i}} \text{。}$

3)宽度

1.2 海上船舶交通流行为特征的拟合

1)灰色理论拟合法[2]

 ${x^{\left( 0 \right)}}\left( i \right) + \rho {H^{\left( 1 \right)}}\left( i \right) = k \text{。}$

 $\frac{{{\rm{d}}{x^{\left( 0 \right)}}}}{{{\rm{d}}x}} + \rho {{\boldsymbol{Y}}}{x^{\left( m \right)}} = {\boldsymbol{B}}{k_1} \text{。}$

 图 2 船舶交通流宽度特征的灰色理论拟合效果 Fig. 2 Fitting effect of grey theory on width characteristics of ship traffic flow

2)高斯拟合法

 $f(x) = A \cdot \exp \left( { - \frac{{{{\left( {x - {\mu _{}}} \right)}^2}}}{{2\sigma _{}^2}}} \right) \text{。}$

 ${\mu _c} = \dfrac{{\displaystyle\sum\limits_W r \cdot {M_{r,c}}}}{{\displaystyle\sum\limits_W {{M_{r,c}}} }} 。$

 ${\sigma _c} = \sqrt {\dfrac{{\displaystyle\sum\limits_{r \in W} {{r^2}} \cdot {M_{r,c}}}}{{\displaystyle\sum\limits_{r \in W} {{M_{r,c}}} }} - \mu _c^2} 。$

 $f(x) = \dfrac{1}{{{\sigma _c}\sqrt {2\text{π} } }} \cdot \exp \left( { - \dfrac{{{{\left( {x - {\mu _c}} \right)}^2}}}{{2\sigma _c^2}}} \right) 。$

 图 3 船舶交通流量的高斯拟合过程示意图 Fig. 3 Schematic diagram of Gaussian fitting process of ship traffic flow
2 基于深度学习算法的船舶交通流行为特征提取系统的开发 2.1 海上船舶交通流行为特征提取系统构成

 图 4 船舶交通流行为特征提取系统示意图 Fig. 4 Schematic diagram of feature extraction system for ship traffic prevalence

1)交通流特征库

2)数据查询

3)用户管理

4)交通流建模

2.2 基于深度学习算法的船舶交通流行为特征提取

 $f(x) = \frac{1}{{1 + {e^{ - bx}}}}\;\;,b > 0 \text{。}$

 图 5 基于深度学习的船舶交通流行为特征提取流程 Fig. 5 Feature extraction process of ship traffic prevalence based on deep learning

 $P\left( X \right) = k\frac{{\left( {\lambda t} \right)k{e^{ - \lambda t}}}}{{k!}}\;\;,\left( {k = 0,1,...} \right) \text{。}$

 $P\left( {X < x} \right) = \sum\limits_{k = 0}^x {} \frac{{\left( {\lambda t} \right)k{e^{ - \lambda t}}}}{{k!}}\; \text{，}$

 $P\left( {x < X < y} \right) = \sum\limits_{k = x}^y k \frac{{\left( {\lambda t} \right)k{e^{ - \lambda t}}}}{{k!}}\; \text{。}$

 图 6 某航线船舶的流量特征仿真曲线 Fig. 6 Flow characteristic simulation curve of ships on a certain route
3 结　语

 [1] 孙瑀隆. 面向宽窄交替航道的船舶交通元胞自动机模型[J/OL]. 系统仿真学报: 1−10. 2022-10-02]. [1] SUN Yu-long. Cellular Automata Model of Ship Traffic for Wide narrow Alternating Channels [J/OL]. Journal of System Simulation: 1−10 [2022-10-2]. [2] 温清洪. 平海湾海上风电场船舶交通流分析[J]. 天津航海, 2022(2): 9-11. WEN Qing-hong. Ship Traffic Flow Analysis of Pinghai Bay Offshore Wind Farm[J]. Tianjin Navigation, 2022(2): 9-11. DOI:10.3969/j.issn.1005-9660.2022.02.003 [3] 廖鹏. 基于GM跟驰模型的内河限制性航道船舶交通流基本图[J]. 东南大学学报(自然科学版), 2022, 52(3): 602-608. LIAO Peng. Basic diagram of ship traffic flow in restricted inland waterway based on GM car following model[J]. Journal of Southeast University (Natural Science Edition), 2022, 52(3): 602-608. DOI:10.3969/j.issn.1001-0505.2022.03.023 [4] 刘钊. 基于二维矩阵分解的船舶交通流预测[J]. 中国航海, 2021, 44(3): 76-83. LIU Zhao. Prediction of ship traffic flow based on two-dimensional matrix decomposition[J]. China Navigation, 2021, 44(3): 76-83. DOI:10.3969/j.issn.1000-4653.2021.03.013