基于粒子群算法的流水线缓冲区容量分配技术研究

发布时间：2018-06-08 00:16

本文选题：粒子群算法 + 缓冲区容量分配　；参考：《兰州理工大学》2017年硕士论文

【摘要】：随着现代市场的发展,制造企业内部以及企业间的相互协作越来越频繁,形成一个连接紧密的网络系统,追求系统整体最优或者共赢成为企业最终赢得生存发展和市场竞争的关键。缓冲区被设计临时用来存储半成品、成品,以降低系统不良状态对其它系统单元乃至整个系统性能的影响或破坏,关系到企业制造系统的整体效能与成败。就现代市场环境而言,无论从供应链物流角度还是企业内部实际生产需要看,均不能被完全取消。以往企业在进行缓冲区容量分配时,由于缺乏成熟理论指导很大程度上依赖于经验,往往造成生产不平衡甚至经营失败的结果。现代企业急需从理论基础上解决在生产要素存在随机事件扰动情况下,如何从系统整体的角度出发对制造系统缓冲区容量进行优化分配的问题,以便达到降低成本和整体最优的目的。基于此,本文针对受随机故障等随机事件影响的流水线系统,深入研究生产线在缓冲区总量固定、生产率最大目标条件下的缓冲区容量优化分配问题。主要的工作内容如下:(1)针对流水线,对分解方法和综合方法等系统性能分析技术进行研究,掌握其思想以及程序算法,通过数值仿真实验了解其优缺点,为后期解决缓冲区容量优化分配问题奠定性能分析技术基础。(2)结合分解方法提出一种改进型多种群粒子群算法的缓冲区容量分配技术,不同于传统仅靠单一种群搜索的进化机制,该算法将粒子群算法引入多个种群中,分别对各种群实行ω线性递减策略,并对各种群附以不同的惯性权重。在与传统粒子群算法的比较中说明该算法的优缺点。(3)以上述研究为基础,针对上述算法收敛速度慢、运算时间长的缺点,提出了一种高斯混沌变异自适应粒子群算法的缓冲区容量分配技术,该算法利用多样性测度函数作为反馈策略,通过平均粒距的引入,使惯性权重随着种群多样性的变化自适应调整,达到较高的收敛速度;并利用高斯混沌变异对算法进行扰动,使其跳出局部最优,从而更好地解决流水线缓冲区容量优化分配技术问题,在与上述算法以及传统算法的比较中验证了该算法的有效性及优越性。(4)最后,结合以上算法,基于MATLAB的GUI图形用户界面开发了“流水线缓冲区容量分配”软件工具,为后期结构更复杂的混杂制造系统的缓冲区容量分配奠定软件模块开发基础。
[Abstract]:With the development of modern market, the cooperation between manufacturing enterprises and enterprises is becoming more and more frequent, forming a closely connected network system. The pursuit of overall optimal or win-win system is the key to win survival and development and market competition. Buffer zones are designed to store semi-finished products temporarily to reduce the impact or destruction of the bad state of the system on the performance of other system units and even the whole system, which is related to the overall efficiency and success of the enterprise manufacturing system. As far as the modern market environment is concerned, it can not be completely cancelled either from the point of view of supply chain logistics or from the point of view of the actual production needs of enterprises. In the past, in the process of buffer capacity allocation, the lack of mature theoretical guidance depended on experience to a great extent, which often resulted in unbalanced production and even failure in operation. Modern enterprises urgently need to solve the problem of how to optimize the buffer capacity allocation of manufacturing system from the point of view of the whole system under the condition of random event disturbance of production factors on the basis of theory. In order to achieve the goal of cost reduction and overall optimization. Based on this, this paper studies the optimal allocation of buffer capacity under the condition of fixed total buffer volume and maximum productivity for pipeline systems affected by random events such as random faults. The main contents of the work are as follows: (1) in view of the pipeline, the system performance analysis techniques, such as decomposition method and synthesis method, are studied, their ideas and program algorithms are grasped, and their advantages and disadvantages are understood through numerical simulation experiments. It lays the foundation of performance analysis for solving the problem of optimizing buffer capacity allocation in the later stage. Combining with decomposition method, a new buffer capacity allocation technique based on improved multi-swarm particle swarm optimization algorithm is proposed. Different from the traditional evolutionary mechanism of single population search, particle swarm optimization (PSO) is introduced into multiple populations, and 蠅 linear decrement strategy is applied to each group, and different inertial weights are attached to each group. In comparison with the traditional particle swarm optimization algorithm, the advantages and disadvantages of the algorithm are illustrated. Based on the above research, the algorithm has the disadvantages of slow convergence and long operation time. A buffer capacity allocation technique for Gao Si chaotic mutation adaptive particle swarm optimization algorithm is proposed. The diversity measure function is used as a feedback strategy and the average particle distance is introduced. The inertial weight can be adjusted adaptively with the variety of population diversity to achieve a higher convergence rate, and the algorithm is disturbed by Gao Si chaos mutation to make it jump out of the local optimum. In order to solve the problem of optimal allocation of pipeline buffer capacity better, compared with the above algorithm and traditional algorithm, the effectiveness and superiority of the algorithm are verified. Finally, the algorithm is combined with the above algorithm. The GUI graphical user interface (GUI) based on MATLAB has developed the "pipeline buffer capacity allocation" software tool, which lays the foundation of the software module development for the buffer capacity allocation of the hybrid manufacturing system with more complex structure in the later stage.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TH186

【参考文献】