AI-Enhanced Generative Design for Efficient Heat Exchangers in Thermoelectric Generators: Revolutionizing Waste Heat Recovery in Thermoelectricity

• Published in: Journal of Mechanical Engineering
• Main Author: Hughes A.R.M.; Singh B.S.B.; Remeli M.F.; Peixer G.F.; Singh W.K.R.
• Format: Article
• Language: English
• Published: UiTM Press 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85215684548&doi=10.24191%2fjmeche.v13i1.2935&partnerID=40&md5=1acd820228613a520c78b324bb49631b

Thermoelectric generators (TEGs) offer the potential for converting waste heat into electricity, but their efficiency, particularly at low temperatures, remains inadequate. Plate-Fin Heat Exchangers (PFHEs) in TEG systems are not fully optimized, resulting in limited efficiency and applicability. The low conversion efficiency of TEGs means only a small fraction of waste heat is utilized, posing challenges to their long-term viability. While Genetic Algorithms (GAs) have shown promise in optimizing heat exchanger designs, advanced methods like Non-dominated Sorting Genetic Algorithm II (NSGA-II) have yet to be fully applied for PFHE TEG design. This study addresses these challenges by using NSGA-II, combined with a semi-empirical model, to optimize PFHE design in TEG systems. The optimization focuses on refining fin design parameters such as number, width, and height while adhering to constraints on fin area and pressure drop. The optimized design achieved a 3.94% increase in output power and a 1.72% increase in efficiency at 373.15 K, with conversion efficiency rising by 154.74% and maximum output power by 549.64% at 428.15 K. In conclusion, this research bridges the gap by applying NSGA-II to enhance PFHE design in TEGs, significantly improving performance and sustainability. Future work will explore alternative models and further optimization to achieve even higher efficiencies. © (2024), (UiTM Press). All rights reserved.