应用遗传算法优化设计微型燃气轮机原表面换热器

Optimized Design of the Original Surface Heat Exchanger of a Micro Gas Turbine by Using the Genetic Algorithm

提出将热力性能设计与遗传算法搜索过程相结合的方法,对适用于100 kW微型燃气轮机的人字形交错波纹板式原表面换热器进行结构优化,分别以重量最轻和换热紧凑度/重量最大作为目标函数,把换热器芯体外形尺寸和换热表面结构尺寸作为待寻求最佳值的优化变量进行优化。遗传算法程序采用二进制编码,锦标赛选择,均匀交叉和单点变异,并采用基于小生境下的共享技术和择优策略。优化结果表明:与原始数据比较,两种目标函数下重量和换热紧凑度都有不同程度的减轻和增加,两侧总压降均比原来略有下降。对比两种目标函数下的优化结果发现,以紧凑度/重量最大为目标函数的优化方法比以重量最轻为目标函数优化方法效果更好。

A method combining the thermal performance design with the genetic algorithm-based searching process was presented for optimizing in structure the herringbone staggered corrugated plate type original surface heat exchanger of a 100 kW micro gas turbine.With the lightest weight and the maximum heat exchange compactness degree/weight serving as the target functions respectively,the overall dimensions of the core and the structural dimensions of the heat exchange surfaces of the heat exchanger in question,employed as the optimization variables of which the optimum values were to be sought,were optimized.The genetic algorithm-based program adopted the binary coding,championship choice,uniform crossover,single point variation,share technology and best choice strategy in a niche.The optimization results show that compared with the original data,the weights and heat exchange compactness degrees under the two target functions decrease and increase respectively to various extents,and the total pressure drops at both sides somewhat decrease compared with the original ones.It has been found through a comparison of the optimization results obtained by using the two target functions that the optimization method with the maximum compactness degree/weight serving as the target function can achieve an effectiveness better than that with the lightest weight serving as the target function.