﻿ 基于群智能优化算法的船舶最优运输路线规划
 舰船科学技术  2024, Vol. 46 Issue (12): 166-169    DOI: 10.3404/j.issn.1672-7649.2024.12.029 PDF

Optimal transportation route planning for ships based on swarm intelligence optimization algorithm
SUN Lin
Tianjin Maritime College, Tianjin 300350, China
Abstract: To avoid potential static obstacle risk areas in the marine transportation environment, and to perform smooth and dynamic obstacle avoidance to reduce the risk of ship collision, a ship optimal transportation route planning method based on swarm intelligence optimization algorithm is studied. This method utilizes the regular hexagonal mesh division method to construct a model of the marine environment in which ships are transported; Establish a transportation route planning model based on swarm intelligence optimization algorithm, and dynamically plan the transportation obstacle avoidance path in the constructed regular hexagonal grid ocean environment. The planning objective is to minimize the navigation distance and the length of the return path during the obstacle avoidance process. Using the Spider Monkey algorithm, solve four planning variables: the time for the ship to navigate to the dynamic obstacle avoidance turning point, the dynamic obstacle avoidance heading change momentum, the time from the dynamic obstacle avoidance turning behavior to the return journey, and the heading change during the return journey, which meet the objective function and constraint conditions. These variables are used as the optimal transportation route planning scheme for ships to achieve the optimal transportation route planning. After testing, it was found that in sea conditions with static and dynamic obstacles, the research method did not pose any collision risks to ships after planning transportation routes, and the routes were smooth.
Key words: swarm intelligence optimization algorithm     ship transportation routes     optimal planning     static obstacles     dynamic obstacles     spider monkey algorithm
0 引　言

1 船舶最优运输路线规划方法 1.1 基于正六边形网格划分的航行环境建模

 \left\{ {\begin{aligned} & {X = X' + \left( {Y' + \left( {Y'\& 1} \right)} \right)/2}，\\ & {Y = Y'} 。\end{aligned}} \right. (1)

 \left\{ {\begin{aligned} & {X' = X - \left( {Y + \left( {Y\& 1} \right)} \right)/2} ，\\ & {Z = - X' - Y} ，\\ & {Y' = Y} 。\end{aligned}} \right. (2)

 图 1 正六边形网格立方体 Fig. 1 Regular hexagonal grid cube
1.2 基于群智能优化算法的运输航线规划模型 1.2.1 最优运输航线规划目标函数设计

 $\begin{split} {{F'}_{s,j}} = &{F_{s,j}} + {\mathrm{rand}}\left( {0,1} \right) \times \left( {{Z_{hj}} - {F_{s,j}}} \right) + \\ & {\mathrm{rand}}\left( { - 1,1} \right) \times \left( {{F_{oj}} - {F_{s,j}}} \right) \\ \end{split}$ (8)

 $I = \left\{ {\begin{array}{*{20}{l}} {\displaystyle\frac{1}{{1 + {q_j}}}}，{{q_j} \geqslant 0} ，\\ {1 + \left| {{q_j}} \right|}，{{q_j} < 0} 。\end{array}} \right.$ (9)

qj使用轮盘赌方法设置蜘蛛猴此阶段被选中的概率，具体为：

 ${q_j} = \frac{{{F_{s,j}}}}{{\displaystyle \sum\limits_{j = 1}^M {{F_{s,j}}} }}。$ (10)

 $\begin{split} {{F'}_{s,j}} =& {F_{s,j}} + {\mathrm{rand}}\left( {0,1} \right) \times \left( {{Z_{Fj}} - {F_{s,j}}} \right) + \\ & {\mathrm{rand}}\left( { - 1,1} \right) \times \left( {{F_{oj}} - {F_{s,j}}} \right) 。\\ \end{split}$ (11)

 $\begin{split} {{F'}_{s,j}} =& {F_{s,j}} + {\mathrm{rand}}\left( {0,1} \right) \times \left( {{Z_{Fj}} - {F_{s,j}}} \right) + \\ & {\mathrm{rand}}\left( {0,1} \right) \times \left( {{F_{s,j}} - {Z_{hj}}} \right)。\\ \end{split}$ (12)

2 实验验证 2.1 实验环境设计

 图 2 船舶运输路线规划系统总体框架 Fig. 2 Overall framework of ship transportation route planning system
2.2 本文方法功能测试