基于遗传算法的锅炉对流受热面优化设计  

Optimal design of convection heating surface of boiler based on genetic algorithm

作　　者：钟崴[1] 吴燕玲[2] 童水光[1] 葛俊旭[2] 周懿[2] 谢金芳[2] 

机构地区：[1]浙江大学热工与动力系统研究所,浙江杭州310027 [2]浙江大学化工机械研究所,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2010年第12期2291-2296,共6页Journal of Zhejiang University：Engineering Science

基　　金：国家"十一五"科技支撑计划资助项目(2006BAF01A46-8)

摘　　要：为优化结构以降低锅炉制造成本,建立以关键结构参数作为决策变量的锅炉对流受热面优化设计模型.提出采用遗传算法进行锅炉对流受热面优化设计的方法,实现完成既定传热任务而减小传热面积的优化目标.选择整数编码形式,并引进变量约束条件、显式函数约束条件和隐式函数约束条件来控制对流受热面结构参数并保证性能指标.利用锅炉设计案例库,采用案例检索与随机生成相结合的方法产生设计方案的初始群体,有效结合设计经验与遗传算法的搜索功能.采用均匀交叉与算术交叉相结合的交叉操作,以及均匀变异产生新个体.采用比例选择与精英保存结合法促进遗传算法向全局最优方向发展.计算结果表明:该方法适用于锅炉对流受热面结构优化设计,能够有效提高设计质量.An optimal design model of boiler convection heating surface which defined key structure parameters as decision variables was established for optimizing the structure and then reducing the manufacturing cost of boiler.An optimization method based on genetic algorithm was proposed to minimize the heat transfer area while accomplish a specified heat transfer task.The structure parameters of convection heating surface were defined by integer-code and controlled by introducing parameter constraints,dominant function constraints and recessive function constraints to satisfy the performance requirements.Supported by boiler design case library,the initial population of design scheme was generated via case-based method combined with random generation method which included design experience and search ability of genetic algorithm.Uniform crossover and arithmetic crossover,as well as uniform mutation were used simultaneously to generate new individuals of design scheme.Proportional selection and elitist preservation were adopted to guide the genetic algorithm process oriented to the global optimization.The results indicate that the model and method is suitable for the structure optimal design of boiler convection heating surface and can improve the design quality effectively.

关 键 词：锅炉 对流受热面 优化设计 遗传算法 

分 类 号：TK212[动力工程及工程热物理—动力机械及工程] TP391.72[自动化与计算机技术—计算机应用技术]