人工蜂群算法的改进及其在经济订货模型中的应用

发布时间：2018-06-16 19:17

本文选题：群体智能算法 + 人工蜂群算法　；参考：《江西理工大学》2017年硕士论文

【摘要】：经济订货批量(Economy Order Quantity,EOQ)是通过平衡各种成本核算使得库存总成本最低的订货量。经济订货批量的计算过程中,需要估计订单的数量以求得更加准确的结果。通过支持向量机(Support Vector Machine,SVM)能够对过往的订单数额进行计算,并预测之后订单数额,进而求得经济订货批量的数值。因此为使得支持向量机的学习效果更加准确,优化支持向量机的方法现已成为热点研究问题之一。人工蜂群算法是一种模拟蜜蜂采蜜行为的群体智能优化算法,由于它具有控制参数少、易于实现、计算简单、鲁棒性强等优点,处理包括优化支持向量机在内的优化问题时有着优异的表现,已被越来越多的研究者所关注。人工蜂群算法主要存在两个缺点:算法特别在处理复杂的优化问题时容易陷入局部最优和过早收敛;算法的探索能力较好,但开发能力不足,收敛速度较慢。本文从多个角度对人工蜂群算法进行改进,提高其在处理复杂优化问题方面的寻优性能,并在此基础上,将算法应用于优化支持向量机以预测经济订货批量模型中的订单预测问题。本文的研究内容主要包括以下两个方面:一方面,为提高算法的优化精度、局部搜索能力,基于现有的名为Bare-bones ABC和HBC的人工蜂群算法的改进算法,提出了一种混合的Bare-bones人工蜂群算法(Hybrid Bare-bones Artificial Bee Colony Algorithm,HBABC)。算法主要改进了以下两个方面:针对算法容易陷入局部最优的方面,引入了HBC算法启发自模拟退火算法的特性对蜜源更新的模型进行了改进;针对算法的收敛性不足的方面,通过启发自Bare-bones ABC的倾向较优个体进行搜索的特性对跟随蜂选择雇佣蜂的方式进行改进。算法通过上述两个改进以提高收敛精度和优化速度。通过使用10个测试函数进行了对比实验,验证了改进算法的有效性。另一方面,本文将HBABC算法用于优化支持向量机的两个参数,并将优化结果用于解决现有的实际问题——基于经济订货批量模型的订单数额及金额的拟合和预测问题。实验结果表明,使用HBABC算法优化的支持向量机得到的拟合和预测结果总体上比使用ABC和BBABC优化的支持向量机表现更加准确。
[Abstract]:Economic order quantity EOQ is the lowest order quantity by balancing all kinds of cost accounting. It is necessary to estimate the quantity of order in order to obtain more accurate results. Through support vector machine support Vector Machine (SVM), we can calculate the amount of order in the past and predict the amount of order after that, and then get the value of economic order batch. Therefore, in order to make the learning effect of SVM more accurate, the method of optimizing SVM has become one of the hot research issues. Artificial bee colony algorithm is a swarm intelligence optimization algorithm to simulate honeybee honey gathering behavior. It has the advantages of less control parameters, easy to implement, simple calculation, strong robustness, and so on. More and more researchers have paid attention to the excellent performance of optimization problems including optimization support vector machines (SVM). The artificial bee colony algorithm has two main disadvantages: the algorithm is prone to fall into local optimum and premature convergence especially when dealing with complex optimization problems, and the algorithm has better exploring ability, but the development ability is insufficient, and the convergence speed is slow. In this paper, the artificial bee colony algorithm is improved from several angles to improve its optimization performance in dealing with complex optimization problems, and on this basis, The algorithm is applied to the optimal support vector machine (SVM) to predict the order forecasting problem in the economic order batch model. The research contents of this paper mainly include the following two aspects: on the one hand, in order to improve the optimization accuracy and local search ability of the algorithm, the improved algorithm based on the existing artificial bee colony algorithm named Bare-bones ABC and HBC is proposed. A hybrid Bare-bones Artificial Bee colony algorithm is proposed. The algorithm mainly improves the following two aspects: to solve the problem that the algorithm is easy to fall into local optimum, the characteristic of self-simulated annealing algorithm inspired by HBC algorithm is introduced to improve the model of honey source update, and the convergence of the algorithm is insufficient. By heuristic Bare-bones ABC, which tends to search better individuals, the method of choosing employment bees is improved. The algorithm improves the convergence accuracy and optimization speed through the above two improvements. The effectiveness of the improved algorithm is verified by 10 test functions. On the other hand, the HBABC algorithm is used to optimize the two parameters of support vector machine, and the optimization results are used to solve the existing practical problems-the fitting and forecasting of the amount and amount of orders based on the economic order batch model. The experimental results show that the performance of SVM optimized by HBABC algorithm is more accurate than that of SVM optimized by ABC and BBABC.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274;TP18

【参考文献】