Genetic Algorithm and Successive Quadratic Programming hybrid method for optimization of shell and tube heat exchangers

Nasrin Sadeghi¹ Saeed Ovaysi²

1) Razi University,
2) Razi University,

Publication : International Congress On Engineering, Technology and Innovation(eticong.com/1st)
Abstract :
An advanced optimization tool can be useful to identify the best and cheapest heat exchanger for a specific duty. In this work, a new hybrid method for optimization of shell and tube heat exchangers is presented. The objective function which is comprised of total annual cost is minimized by varying three optimization variables; shell diameter, tubes diameter, and baffle spacing. To overcome the high cost and unreliability associated with Genetic Algorithm (GA), in the present work a hybrid of GA and Successive Quadratic Programming (SQP) is used to obtain the optimal design of several heat exchangers already studied. In all cases, a hybrid method improved design with lower total annual costs (Ctotal) was achieved, e.g. the calculated Ctotal by GA is 36097 which is reduced to 23302 by a hybrid method. The results indicated enhanced accuracy and lower computational load of the proposed method compared to those of pure GA.
Keywords : Shell and tube heat exchanger Genetic algorithm Successive Quadratic Programming Hybrid algorithm Optimization