Non-dominated Sorting Genetic Algorithm II and Particle Swarm Optimization for design optimization of Shell and Tube Heat Exchanger

Abstract

Optimization methods are applied in Shell and Tube Heat Exchanger (STHE) design to reduce the cost of the device. Various existing optimization techniques such as Particle Swarm Optimization (PSO), Adaptive Range Genetic Algorithm (ARGA) are applied in the design of STHE. Existing optimization methods used in STHE design, have the limitation of lower convergence and easily trap into local optima. In this research, the hybrid method of Non-dominated Sorting Genetic Algorithm II (NSGA II) and PSO method is proposed to reduce the cost in STHE design. The NSGA II method is applied to improve the exploration and PSO method is applied to improve exploitation of search process. The hybrid method has objective function of total cost and overall heat transfer of the model to improve the performance. The NSGA II has strong exploration in the search due to the non-dominated search process and also provides good convergence. The PSO method is applied in the best solution of NSGA II and the PSO method has the advantage of strong exploitation that escapes from the local optima. The hybrid NSGA II-PSO method is tested on three test cases and is compared with existing optimization methods to analyze its performance. The result shows that the hybrid NSGA II-PSO method has a 4.85% lesser total cost in case 1 and 1.51% lesser total cost in case 2, when compared to the ARGA method.