基于群体智能优化理论的辐射换热腔体几何反设计

发布时间：2017-12-28 13:05

本文关键词：基于群体智能优化理论的辐射换热腔体几何反设计　出处：《哈尔滨工业大学》2015年硕士论文　论文类型：学位论文

【摘要】：辐射换热设备广泛存在于工业锅炉、红外反射炉、材料加工成型等多种工业领域,腔体设计的合理与否直接影响到安全问题。几何反设计是近些年新兴的一种设计方法,基本思路是首先根据设计要求建立相应的目标函数,然后采用优化方法对其优化,最终达到设计目的。反设计方法具有过程简单、设计周期短、求解精度高等优点,因而得到越来越广泛的关注和应用。目前辐射腔体的几何反设计多采用基于对目标函数求导的方法,这类方法具有收敛速度快、结果稳定性好等优点,但其梯度求解过程繁琐复杂,而且对初值依赖性较大。群体智能优化算法随机地从某一个解出发,然后根据相应算法的搜寻原则,寻找下一个使目标函数更小的解,逐步搜索直至得到最优解,这种求解方式逻辑简单,易于编程实现,而且算法本身具有并行性、自组织性、灵活性等优点,因而越来越得到人们的重视。本文基于群体智能优化理论求解了二维辐射换热腔体的几何反设计问题,主要研究内容包括以下几个方面:(1)采用贴体坐标系下离散坐标法求解不规则几何形状介质内的辐射换热问题,并与有限体积法计算结果进行对比,验证正问题模型计算程序的可靠性,根据求得的边界热流密度建立相应的目标函数。(2)概述群体智能优化算法的主要特点及其分类,简单介绍群搜索方法、人工鱼群算法、蚁群优化算法以及蛙跳算法等四种常见群体智能优化算法的求解基本思路,并重点阐述磷虾群算法和微粒群优化算法的基本原理及其求解过程,为后续几何反设计研究提供理论支撑。(3)分别采用磷虾群算法、标准微粒群优化算法和随机微粒群优化算法求解辐射换热腔体几何反设计问题,通过几何优化,达到设计表面指定区域热流密度均匀分布的设计目的,并比较了三种求解方法在辐射腔体几何反设计问题中的计算性能;分析了几何反设计中设计表面控制点数量、介质物性及边界黑度等参数对设计结果的影响。
[Abstract]:Radiant heat exchanger widely exists in many industrial fields, such as industrial boiler, infrared reflector, material processing and molding. Whether the cavity design is reasonable or not directly affects the safety problem. Geometric inverse design is a new design method in recent years. The basic idea is to set up the corresponding objective function according to the design requirements, and then use the optimization method to optimize it, and ultimately achieve the purpose of design. The anti design method has the advantages of simple process, short period of design and high precision of solving, so it has been paid more and more attention and application. At present, the geometric inverse design of radiation cavities is based on the derivation method of objective function. This method has the advantages of fast convergence and good stability, but its gradient solution is complicated and complex, and it depends on initial value. Swarm intelligence optimization algorithm randomly from the point of view of a solution, and then according to the principle of the corresponding search algorithm, to find a better solution to search until the optimal solution, this solution logic is simple, easy in programming, and the algorithm has the advantages of self organization, parallelism and flexibility therefore, more and more people's attention. The geometry of swarm intelligence optimization theory to solve the two-dimensional heat radiation cavity based on inverse design problems, the main research contents include the following aspects: (1) using body fitted coordinates discrete coordinate method of irregular geometry solving radiation heat transfer problems in medium, and the results are compared with the finite volume method, reliability calculation program verification is the problem of model, according to heat flux boundary obtained the corresponding objective function is established. (2) an overview of swarm intelligence optimization algorithm the main characteristics and classification, introduces the swarm search method, artificial fish swarm algorithm, ant colony optimization algorithm and SFLA four kinds of swarm intelligence optimization algorithm for solving the basic ideas, and focuses on the krill swarm algorithm and particle swarm optimization algorithm is the basic principle and solving process, theory support for the subsequent research of inverse geometry design. (3) using krill swarm algorithm and standard particle swarm optimization algorithm and stochastic particle swarm optimization algorithm for solving radiative heat transfer cavity geometry inverse design problem, the geometric optimization, achieve the goal of the design design of surface heat flux distribution of the designated area, and compare the three method calculation of performance in the radiation cavity inverse geometry design problems the analysis of the influence of inverse geometry design; design of surface control point number, dielectric properties and boundary emissivity parameters on the design results.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TK172

【参考文献】