摘要
介绍了一种由两级管壳式换热器组成的换热器系统的优化设计数学模型,数学模型属于典型的有约束非线性规划问题。目前对有约束非线性规划问题还没有通用的求全局最优解的算法。转轴直径搜索可行方向法(DSFD方法)是一种比较有效的求解有约束非线性规划问题的算法,但它只能得到局部最优解。将模拟退火(SimulatedAnnealing)算法结合DSFD算法,构成了一种DSFD-SA-DSFD算法。在应用模拟退火算法的同时引入了罚函数法,将有约束非线性问题转化为无约束非线性问题。计算结果表明,DSFD-SA-DSFD算法能较快得到换热器优化问题的最优解,克服了单纯用DSFD算法只能得到局部最优解和单纯用SA算法效率不高的缺点。
This paper introduces a kind of optimal design mathematic model which handles the heat exchanger system consisting of two stages shell-and-tube heat exchanger. This model belongs to the typical constrained non-linear programming problem. Currently, there is no general algorithm to find the global optimal solution for the constrained non-linear programming problems. Direct search feasible direction (DSFD) is an effective algorithm for solving the constrained non-linear programming problem, however, it could only find the local optimal solution. This paper integrates the simulated annealing (SA)algorithm and DSFD algorithm to constitute a new algorithm called DSFD-SA-DSFD. Simultaneously, this paper integrates the SUMT (the sequential unconstrained minimization technique) into SA algorithm and converts the constrained problem to unconstrained problem. The result shows that, DSFD-SA-DSFD algorithm could find the optimal solution for the problem of heat exchanger optimization quickly, not only overcome the limit of the DSFD algorithm which could only find the local optimal solution, but also solve the inefficiency of SA algorithm. Figs 3, tables 11 and refs 8.
出处
《动力工程》
CSCD
北大核心
2004年第2期285-290,共6页
Power Engineering
关键词
动力机械工程
换热器
优化设计
模拟退火
罚函数
power and mechanical engineering
heat exchanger
optimal design
simulated annealing
SUMT (the sequential unconstrained minimization technique)
