模拟追逐算法及其应用研究
发布时间：2018-07-23 16:20
【摘要】：智能优化算法是通过模拟自然生物进化或者社会行为而提出的一种随机搜索算法.这些算法具有原理简单、模型易于实现的特征,能够解决许多传统方法难以解决的复杂问题,被广泛应用于各个领域.本文借鉴长跑比赛运动员追逐竞赛的过程,建立了数学模型并应用于优化问题的求解,提出了一种新的智能优化算法——模拟追逐算法(Simulated Pursuit Algorithm).模拟追逐算法融合了试探性开拓与有目的性的追逐相结合的优点,是一种有效的群体全局优化算法.本文首先介绍基本模拟追逐算法,分析了算法的基本原理;为提高算法种群多样性,引入协作算子,提出了协同进化的模拟追逐算法;为将模拟追逐算法应用从连续优化问题推广到离散优化问题,对探测算子与追逐算子给出了新的设计定义,提出一种改进的模拟追逐算法求解TSP问题.具体研究工作如下三个方面:1、提出了新的群体智能算法——模拟追逐算法.该算法设计了追逐算子与探测算子;领先个体执行探测算子操作以便获得更优位置,落后个体为取得竞争优势,设定追赶目标,执行追逐算子操作,完成跟随超越,从而实现群体进化寻优.对追逐算子进行性能分析;采用六个典型的测试函数进行仿真实验,分析了算法的求解精度、收敛速度以及稳定性.仿真实验表明,模拟追逐算法有较快的收敛速度和较高的求解精度,是一种稳定的优化算法.2、为了维持算法搜索过程中种群多样性,本文在基本的模拟追逐算法(SPA)中引入三种协作算子,充分共享个体间的信息,提出一种协同进化模拟追逐算法.四个基准函数测试表明,在算法后期加入翻转交叉算子可避免产生过多的重复解;改进后的算法平衡了搜索的集中性和种群多样性,提高算法跳出局部最优的能力;协同进化模拟追逐算法在寻优能力与收敛速度均优于基本的模拟追逐算法.3、为将模拟追逐算法推广应用到组合优化问题,本文提出了一种改进的模拟追逐算法求解TSP.算法采用贪心策略与对称策略初始化种群;定义交换运算、交换矩阵,对追逐算子和探测算子给出新的设计定义;仿真实验表明,改进的模拟追逐算法对TSP有更高的求解精度,是一种有效算法.模拟追逐算法应用于求解TSP问题的研究,提供了将模拟追逐算法运应用于处理离散优化问题的模板,扩宽了模拟追逐算法的应用领域.
[Abstract]:Intelligent optimization algorithm is a random search algorithm proposed by simulating natural evolution or social behavior. These algorithms have the characteristics of simple principle, easy to implement the model, and can solve many complex problems which are difficult to be solved by traditional methods. They are widely used in various fields. In this paper, using the process of long distance race for reference, a mathematical model is established and applied to solve the optimization problem. A new intelligent optimization algorithm, simulation chase algorithm (Simulated Pursuit Algorithm)., is proposed in this paper. Simulation chase algorithm, which combines the advantages of exploratory exploration and purposeful pursuit, is an effective group global optimization algorithm. In this paper, the basic simulation chase algorithm is introduced, and the basic principle of the algorithm is analyzed, in order to improve the diversity of the algorithm population, the cooperative operator is introduced, and the simulation chase algorithm of co-evolution is proposed. In order to extend the application of simulation chase algorithm from continuous optimization problem to discrete optimization problem, a new design definition of detection operator and chase operator is given, and an improved simulation chase algorithm is proposed to solve the TSP problem. The specific research work is as follows: 1. A new swarm intelligence algorithm-simulation chase algorithm is proposed. In this algorithm, chase operator and detection operator are designed, leading individual performs detection operator operation in order to obtain better position, backward individual sets chase target, executes chase operator operation, and accomplishes following surpassing in order to gain competitive advantage. Thus, the optimization of population evolution can be realized. The performance of the chase operator is analyzed, and six typical test functions are used to simulate the algorithm. The accuracy, convergence speed and stability of the algorithm are analyzed. The simulation results show that the simulation chase algorithm has faster convergence speed and higher accuracy, and is a stable optimization algorithm. In order to maintain the diversity of the population in the search process of the algorithm, In this paper, three kinds of cooperative operators are introduced into the basic simulation chase algorithm (SPA) to fully share the information between individuals, and a co-evolution simulation chasing algorithm is proposed. The test of four benchmark functions shows that adding overturning crossover operator in the later stage of the algorithm can avoid too many repetitive solutions and the improved algorithm balances the search centrality and population diversity and improves the ability of the algorithm to jump out of the local optimum. In order to extend the simulation chase algorithm to combinatorial optimization problem, an improved simulation chase algorithm is proposed to solve the TSPs in order to extend the simulation chase algorithm to the combinatorial optimization problem. The algorithm uses greedy strategy and symmetric strategy to initialize the population, define exchange operation, exchange matrix, give a new design definition of chase operator and detection operator, and the simulation results show that the improved simulation chase algorithm has higher accuracy for TSP. Is an effective algorithm. The simulation chase algorithm is applied to the study of solving the TSP problem. It provides a template for the simulation chase algorithm to deal with discrete optimization problems and widens the application field of the simulation chase algorithm.
【学位授予单位】：广东工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18
本文编号：2139956
[Abstract]:Intelligent optimization algorithm is a random search algorithm proposed by simulating natural evolution or social behavior. These algorithms have the characteristics of simple principle, easy to implement the model, and can solve many complex problems which are difficult to be solved by traditional methods. They are widely used in various fields. In this paper, using the process of long distance race for reference, a mathematical model is established and applied to solve the optimization problem. A new intelligent optimization algorithm, simulation chase algorithm (Simulated Pursuit Algorithm)., is proposed in this paper. Simulation chase algorithm, which combines the advantages of exploratory exploration and purposeful pursuit, is an effective group global optimization algorithm. In this paper, the basic simulation chase algorithm is introduced, and the basic principle of the algorithm is analyzed, in order to improve the diversity of the algorithm population, the cooperative operator is introduced, and the simulation chase algorithm of co-evolution is proposed. In order to extend the application of simulation chase algorithm from continuous optimization problem to discrete optimization problem, a new design definition of detection operator and chase operator is given, and an improved simulation chase algorithm is proposed to solve the TSP problem. The specific research work is as follows: 1. A new swarm intelligence algorithm-simulation chase algorithm is proposed. In this algorithm, chase operator and detection operator are designed, leading individual performs detection operator operation in order to obtain better position, backward individual sets chase target, executes chase operator operation, and accomplishes following surpassing in order to gain competitive advantage. Thus, the optimization of population evolution can be realized. The performance of the chase operator is analyzed, and six typical test functions are used to simulate the algorithm. The accuracy, convergence speed and stability of the algorithm are analyzed. The simulation results show that the simulation chase algorithm has faster convergence speed and higher accuracy, and is a stable optimization algorithm. In order to maintain the diversity of the population in the search process of the algorithm, In this paper, three kinds of cooperative operators are introduced into the basic simulation chase algorithm (SPA) to fully share the information between individuals, and a co-evolution simulation chasing algorithm is proposed. The test of four benchmark functions shows that adding overturning crossover operator in the later stage of the algorithm can avoid too many repetitive solutions and the improved algorithm balances the search centrality and population diversity and improves the ability of the algorithm to jump out of the local optimum. In order to extend the simulation chase algorithm to combinatorial optimization problem, an improved simulation chase algorithm is proposed to solve the TSPs in order to extend the simulation chase algorithm to the combinatorial optimization problem. The algorithm uses greedy strategy and symmetric strategy to initialize the population, define exchange operation, exchange matrix, give a new design definition of chase operator and detection operator, and the simulation results show that the improved simulation chase algorithm has higher accuracy for TSP. Is an effective algorithm. The simulation chase algorithm is applied to the study of solving the TSP problem. It provides a template for the simulation chase algorithm to deal with discrete optimization problems and widens the application field of the simulation chase algorithm.
【学位授予单位】：广东工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18
【参考文献】
相关期刊论文 前10条
1 Andreea Koreanschi;Oliviu Sugar Gabor;Joran Acotto;Guillaume Brianchon;Gregoire Portier;Ruxandra Mihaela Botez;Mahmoud Mamou;Youssef Mebarki;;Optimization and design of an aircraft's morphing wing-tip demonstrator for drag reduction at low speed, Part Ⅰ Aerodynamic optimization using genetic, bee colony and gradient descent algorithms[J];Chinese Journal of Aeronautics;2017年01期
2 周向华;杨侃;王笑宇;;狼群算法在水电站负荷优化分配中的应用[J];水力发电;2017年02期
3 张超;李擎;王伟乾;陈鹏;冯毅南;;基于自适应搜索的免疫粒子群算法[J];工程科学学报;2017年01期
4 易云飞;林晓东;蔡永乐;;求解旅行商问题的改进粒子群算法[J];计算机工程与设计;2016年08期
5 刘伟;谢月珊;;模拟追逐算法[J];广东工业大学学报;2016年02期
6 王建群;贾洋洋;肖庆元;;狼群算法在水电站水库优化调度中的应用[J];水利水电科技进展;2015年03期
7 侯淑静;;求解TSP问题的几种算法比较[J];黄冈职业技术学院学报;2015年01期
8 雷玉梅;;基于改进遗传算法的大规模TSP问题求解方案[J];计算机与现代化;2015年02期
9 吴虎胜;张凤鸣;战仁军;李浩;梁晓龙;;利用改进的二进制狼群算法求解多维背包问题[J];系统工程与电子技术;2015年05期
10 李阳;李文芳;马骊;樊锁海;;混合退火算法求解旅行商问题[J];计算机应用;2014年S1期
相关硕士学位论文 前2条
1 张鹏;多蜂群协同进化算法及其应用研究[D];山东师范大学;2014年
2 张雅妮;基于粒子群算法模糊控制自动舵的研究与仿真[D];哈尔滨工程大学;2010年
，本文编号：2139956
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2139956.html
