鸽群优化算法及其应用研究

发布时间：2018-10-25 09:20

【摘要】：鸽群优化算法是新的启发式算法,是由段海滨教授等人于2014年首次提出。鸽群算法的思想是模拟鸽群利用地球磁场和地标组合来归巢的过程。鸽群算法具有原理相对简单、所需调整的参数较少、比较容易实现等特点。还有计算相对简单,鲁棒性较强等显著优势,相对于其他部分算法而言还有收敛速度较快的优势。与此同时鸽群优化算法还存在不足之处,该算法有收敛精度偏低,容易出现局部最优的情况,稳定性较差等缺点。因此,鸽群优化算法在理论方面和应用方面,是有待于更深入的研究和更广泛的扩展。本文针对鸽群优化算法所存在的不足进行了分析,从添加收敛因子、位置因子、速度因子和子群变异策略等方面对鸽群算法进行了改进,还将改进后的算法应用到实际的优化问题当中。所涉及的主要工作内容将总结为以下3个方面:(1)采用添加收敛因子、增加位置因子和速度因子等策略对鸽群算法进行改进。不仅增强鸽子的飞行活力,提高鸽子种群的多样性,并且能够有效的避免鸽群的早熟收敛现象,使鸽群优化算法具有一定的竞争力。并且完成了改进后的鸽群优化算法的相关标准函数优化的测试。(2)通过添加子群变异策略对鸽群优化算法进行改进,将子群变异策略的思想应用到鸽群优化算法中,克服了鸽群优化算法容易早熟收敛情况,还增大了鸽子种群潜在的搜索空间。为了增强鸽群优化算法的局部搜索能力,还对贪心策略进行引入,并且将改进后的鸽群优化算法应用于0-1背包问题的求解。(3)通过把鸽群优化算法和模拟退火算法进行互补融合。结合后的算法不仅仅具有鸽群算法的特点,还可以根据概率性进行劣向的转移,并且以一定的概率去接受劣解,可以让鸽群优化算法跳出局部最优解的情况,从而达到实现全局最优的目的。在与算法融合的基础上还对鸽群优化算法引入自适应温度衰变系数,可以根据当前的环境自动调整搜索条件,来达到提高搜索效率的目的。本章还将改进后的鸽群算法应用于求解无人潜水器路径规划的问题当中,以此来增加改进后的鸽群算法的应用范围,同时也表明了算法的有效性和可行性。
[Abstract]:Pigeon swarm optimization, a new heuristic algorithm, was first proposed by Professor Duan Haibin in 2014. The idea of pigeon swarm algorithm is to simulate the homing process of pigeon swarm using the combination of geomagnetic field and landmarks. Pigeon swarm algorithm has the advantages of relatively simple principle, few parameters to be adjusted and easy to implement. There are also obvious advantages such as relatively simple computation and strong robustness, and the advantages of faster convergence rate compared with other algorithms. At the same time, the pigeon swarm optimization algorithm has some shortcomings, such as low convergence accuracy, local optimum and poor stability. Therefore, pigeon swarm optimization algorithm needs to be further studied and extended in theory and application. This paper analyzes the shortcomings of pigeon swarm optimization algorithm and improves the algorithm from the aspects of adding convergence factor, position factor, speed factor and subgroup mutation strategy. The improved algorithm is also applied to practical optimization problems. The main work involved will be summarized as follows: (1) improving pigeon swarm algorithm by adding convergence factor, increasing position factor and speed factor. It can not only enhance the flight vitality of pigeons, improve the diversity of pigeon population, but also effectively avoid the phenomenon of premature convergence of pigeon population, so that the pigeon swarm optimization algorithm has certain competitiveness. And completed the improved pigeon swarm optimization algorithm related standard function optimization test. (2) by adding subgroup mutation strategy to improve the pigeon swarm optimization algorithm, the idea of subgroup mutation strategy is applied to pigeon swarm optimization algorithm. It overcomes the premature convergence of pigeon swarm optimization algorithm and increases the potential search space of pigeon population. In order to enhance the local search ability of pigeon swarm optimization algorithm, the greedy strategy is also introduced, and the improved pigeon swarm optimization algorithm is applied to solve the 0-1 knapsack problem. (3) the pigeon swarm optimization algorithm is combined with simulated annealing algorithm to solve the 0-1 knapsack problem. The combined algorithm not only has the characteristics of pigeon swarm algorithm, but also can transfer the bad solution according to the probability, and accept the inferior solution with a certain probability, so that the pigeon swarm optimization algorithm can jump out of the local optimal solution. In order to achieve the goal of global optimization. Based on the fusion with the algorithm, the adaptive temperature decay coefficient is introduced to the pigeon swarm optimization algorithm, which can automatically adjust the search conditions according to the current environment to achieve the purpose of improving the search efficiency. In this chapter, the improved pigeon swarm algorithm is applied to solve the path planning problem of unmanned submersible vehicle, so as to increase the scope of application of the improved pigeon swarm algorithm, and also show the effectiveness and feasibility of the algorithm.
【学位授予单位】：广西民族大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】