With the discovery of promising materials, the optimisation of the design and selection of thermoelectric coolers (TEC) can further advance their engineering applications in various fields. Thus, this study proposed two parameters: the device thermal resistance and linear current density of the TEC, to build a direct link between the design, selection, and operation of the TEC. Further, a surrogate model in terms of support vector regression was constructed for TEC to reduce the computing complexity, resources, and time. Thus, an optimal design and selection strategy for TEC was developed for practical multiple application scenarios. Consequently, the influences of the optimisation objective, design electrical current, and hot- and cold-side heat exchangers on the design results were discussed. A case study was conducted to demonstrate the usage method and validate the effectiveness of the proposed strategy. The design, procurement, and operating costs were analysed. The proposed strategy directly transformed design results into selection results with a deviation of <11 %. Further, using the proposed strategy, the design cost was almost negligible, and the procurement and operating costs were reduced by 74.6 % and 46.43 %, respectively. It was found that a designed TEC with a smaller device thermal resistance and higher linear current density could exhibit a higher cooling capacity; otherwise, it could obtain a higher coefficient of performance (COP). Comprehensive cooling performance of TEC designed for multiple working conditions decreased by only 6.3 % under different working conditions. This study bridges the gap between the designing and selection of TEC and helps users choose more efficient and suitable TECs for practical applications, which may further promote the widespread use of TEC.
Read full abstract