本科毕设论文--基于邻域克隆选择学习算法的分馏系统资源优化.docVIP

题目： （中文） 基于邻域-克隆选择学习算法的分馏系统资源优化 （英文） Optimization of distillation resources based on neighborhood- clonal selection learning algorithm 诚 信 承 诺 我谨在此承诺：本人所写的毕业论文《基于邻域-克隆选择学习算法的分馏系统资源优化》均系本人独立完成，没有抄袭行为，凡涉及其他作者的观点和材料，均作了注释，若有不实，后果由本人承担。 承诺人（签名）： 年 月 日 摘 要 【摘要】针对化工分馏系统资源优化问题，本文在结合生物免疫的克隆选择和人工免疫网络算法的基础上，采用了Agent的思想，提出了一种邻域-克隆选择学习的全局优化算法（N- Clonalg）。N- Clonalg通过定义网格化的邻域操作环境，使每个个体完成全局-局部的搜索操作。不同于其它人工免疫算法，N-Clonalg的主要搜索算子有邻域克隆选择、邻域竞争和自学习算子，能有机结合全局与局部搜索，多峰测试函数表明能较好地克服了克隆选择算法（Clonalg）的早熟及人工免疫网络算法（Opt-aiNet）收敛速度慢现象，能在个体空间中有效地实现全局最优搜索。分馏系统资源优化的实例表明它具有较好的最优解搜索性能，能较好地实现化工中的寻优问题。但是资源优化模型运行起来速度较慢，针对这一现象，结合神经网络的特点，先用rbf神经网络逼近化工模型，当逼近到一定程度，再调用化工模型进行资源优化。利用这一思维，更好地进行了分馏系统资源优化的进行，更加快捷有效。 【关键词】克隆选择学习；邻域-克隆选择学习算法；多模态优化；分馏装置；RBF神经网络。 Optimization of distillation resources based on neighborhood- clonal selection learning algorithm Abstract 【ABSTRACT】Faced with the problem of resources optimization in chemical fractionation system, in this paper, we propose a global optimization algorithm(N-Clonalg) of neighbourhood - clonal selection study, which is combined with biological immune clonal selection and artificial immune algorithm based on the network and adapts the idea of Agent. By defining the grid operating environment of the neighbourhood, so N-Clonalg make each individual complete search operations from local to global. Unlike other artificial immune algorithm, N-Clonalg has the main search operators, such as the neighbourhood clonal selection, neighbourhood competition and self-learning operator.Combining global and local search organic, N-Clonalg overcomes the phenonmena of Clonalg precocious and Opt-aiNet slow convergence better and achieves global optimal search effectively in the individual space, which is proved in Multi-peak function test. Indeed, fractional system resource optimization examples also show that it has better search performance of the optimal