多种群混合蛙跳算法在集装箱堆场场桥路径规划中的应用

发布时间：2018-02-08 13:54

本文关键词： 蛙跳算法场桥路径规划遗传算法集装箱堆场　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：全球经济的飞速发展使得贸易运输业务急剧增长。航运在贸易运输中占有重要地位,其中集装箱运输是航运中最主要的运输方式之一。集装箱码头为了提高经济效益,需要提高自身的工作效率。而集装箱堆场场桥的工作效率是影响码头整体效率的关键问题之一。解决好场桥路径规划问题,可有效提高集装箱码头的作业效率,提升码头整体效益,从而提高港口的竞争力。本文采用群智能算法求解集装箱堆场场桥的路径规划问题。在已知集装箱堆场堆放状况下,根据给定的提箱任务,首先建立了以场桥移动距离最短为目标的单台和多台场桥的数学优化模型。然后针对该离散组合优化问题,研究了一种高效的求解算法。基于较为新颖的蛙跳算法,对其存在的缺陷加以改进,提出了多种群混合蛙跳算法。其基本思想是采用并行策略,将整个蛙群分为三个子群体,它们分别侧重于向全局最优学习以加快收敛速度,和在较优个体附近的局部开发以及全局搜索,维持群体多样性和防早熟。三者定期进行信息交换,以发挥各自所长,优势互补,提高算法整体性能。此外,基于遗传算法的交叉和变异算子的引入能使所提算法适用于求解此类路径规划等离散组合优化问题;而与模拟退火思想的混合能够改善蛙跳算法对最优个体附近局部搜索能力的不足,可望进一步加速收敛且有利于防止早熟。为了验证所提算法的性能,文中将其应用于求解已知最优解的经典函数优化和旅行商问题,优化结果验证了其可行性和有效性。在此基础上,进一步以集装箱堆场场桥路径规划问题为工程背景,针对所提算法给出了其具体实现的编码方法,以及交叉和变异策略。将所提算法应用于前述数学优化模型之中,进行仿真与测试,分别求解了针对两个实例的单台场桥和多台场桥的路径规划问题,并对结果进行了分析和对比。研究表明,提出的算法对于该路径规划问题是有效的,获得了较好的优化结果,所得的场桥路径规划方案,工程适用且令人满意。本文的工作能够为集装箱堆场场桥的实际作业操作提供参考和借鉴,以达到提高码头的作业效率和经济效益的目的。研究具有一定的理论意义和实用价值。
[Abstract]:With the rapid development of the global economy, trade and transportation business is growing rapidly. Shipping plays an important role in trade transportation, among which container transportation is one of the most important transportation modes. The efficiency of the container yard bridge is one of the key problems affecting the overall efficiency of the terminal. To solve the problem of the route planning of the yard bridge can effectively improve the operational efficiency of the container terminal. In this paper, we use swarm intelligence algorithm to solve the path planning problem of container yard bridge. Under the condition of known container yard stacking, according to the given container task, Firstly, the mathematical optimization models of single and multiple field bridges with the shortest moving distance are established. Then, an efficient algorithm for solving the discrete combinatorial optimization problem is proposed, which is based on a novel leapfrog algorithm. In order to improve its shortcomings, a multi-group hybrid leapfrog algorithm is proposed. The basic idea is to divide the whole frog population into three subpopulations by using parallel strategy, which respectively focus on learning to the global optimum to speed up the convergence rate. And local development in the vicinity of superior individuals and global search to maintain population diversity and prevent precocity. The three regularly exchange information to give play to their respective strengths, complement each other, and improve the overall performance of the algorithm. The introduction of crossover and mutation operators based on genetic algorithm can make the proposed algorithm suitable for solving such discrete combinatorial optimization problems as path planning. Mixing with simulated annealing can improve the local search ability of the leapfrog algorithm near the optimal individual, which is expected to accelerate convergence and prevent precocity. In order to verify the performance of the proposed algorithm, In this paper, it is applied to the classical function optimization and traveling salesman problem of known optimal solution. The optimization results verify its feasibility and effectiveness. On this basis, the engineering background of the bridge path planning problem of container yard is further taken as the engineering background. The coding method, crossover and mutation strategy of the proposed algorithm are given. The proposed algorithm is applied to the above mathematical optimization model for simulation and test. The path planning problem of single field bridge and multi field bridge with two examples is solved, and the results are analyzed and compared. The results show that the proposed algorithm is effective for the path planning problem, and good optimization results are obtained. The project is applicable and satisfactory. The work in this paper can provide reference and reference for the practical operation of the container yard bridge. In order to improve the efficiency and economic benefit of the wharf, the research has certain theoretical significance and practical value.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U691.3;TP18

【参考文献】