遗传-模拟退火算法优化设计管壳式换热器

Optimization of a shell-and-tube heat exchanger based on a genetic simulated annealing algorithm

依据Bell-Delaware法对壳程流体进行压降和传热的计算,选择管径、管长、折流挡板数等结构参数作为主要设计变量,参考了美国管式换热器制造商协会(Tubular Exchanger Manufacturers Association,TEMA)标准作为相关约束条件,以换热器的年度总费用最低为目标函数,建立了管壳式换热器优化设计数学模型,并基于遗传-模拟退火算法(GA-SA)进行求解。文献算例的对比结果表明:算法能较好地权衡换热器的换热面积费用和泵的操作费用并搜索到全局最优解,从而获得总费用较低的换热器主要结构参数。针对一个实际工程项目,考虑换热器设计裕度要求,计算结果与商业化软件HTRI的预测值接近,说明所设计的换热器实际可行。同时克服了HTRI需要设计者的经验确定设计变量和无法保证经济性最优的不足。

A mathematical model was developed to optimize the design of a shell-and-tube heat exchanger based on design data obtained by using the Bell-Delaware method to describe the pressure drop and heat trans{er on the shell-side. The design variables were the tube diameter, the tube length, and other geometric parameters with the Tubular Exchanger Manufacturers Association （TEMA） standard taken as the reference for the constraints and the minimum total heat exchanger cost as the objective. The solution used the genetic simulated annealing algorithm （GA-SA）. This method more effectively balances the heat exchanger area cost and pumping cost than previous methods by searching for the global optimal solution for the main geometric heat exchanger parameters with the minimum total cost. With the margin requirement for heat exchanger designs for specific industrial projects, these results are close to those given by commercial HTRI software, which indicates that this heat exchanger design method is reliable. This method guarantees the economic optimum without an empirical method to optimize the design variables in the heat exchanger design which is a major weakness of HTRI software packages.