Multi-objective optimization of thermo-acoustic devices using teaching-learning-based optimization algorithm

Abstract

The current article presents three case studies based on a multi-objective optimization approach to optimize the performance of thermo-acoustic devices by obtaining the best possible set of geometrical characteristic parameters. In case study 1, the performance of a thermo-acoustic refrigerator is measured in terms of three objectives namely, acoustic cooling load (ΦC), coefficient of performance, and acoustic power loss (W02). Each objective is assigned a weight to facilitate suitable user-defined significance. The case study 2 aims to optimize a thermoacoustic prime-mover. The influence of stack position and its length, resonator length, plate thickness, and plate spacing are considered as design variables. Two objectives namely, pressure amplitude (P) and frequency (f) are considered as objectives for multi-objective optimization of the thermo-acoustic prime-mover. In case study 3, the performance of a thermo-acoustic engine is measured in terms of five objectives namely, work output (W), viscous loss (Rv), conductive heat loss (Qcond), convective heat loss (Qconv), and radiative heat loss (Qrad). Since the multiple objectives are to be optimized simultaneously, each objective is assigned a weight to facilitate suitable user-defined significance. The multi-objective optimization is carried out by the teaching-learning-based optimization algorithm. The results of application of teaching-learning-based optimization algorithm are compared with the results of General Algebraic Modeling System, response surface methodology, and experimental results. The results of the teaching-learning-based optimization algorithm are found better compared to those given by the other approaches.