Abstract Using thermoelectric power generation system is one of the clean devices for converting heat to electrical power which is applied to recover waste energy. In this study, paying special attention to the investment and second law efficiency, a TEG system equipped with hot and cold heat-exchangers is considered and optimized for an automobile exhaust. First, the P-N couples is discretized and divided to a specified number of elements. By applying the governing equations to each element and considering the condition of the flow regime of the heat-exchangers, an enhanced numerical algorithm is developed to analyze the system and calculate the objective functions including power output, second law efficiency, investment and a compound function which is a sum of power, second law efficiency and the investment. Length of the legs, number of the P-N couples, hydraulic diameter of the heat exchangers and the hot fluid mass flow rate are considered as decision variables. Single and multi-objective optimization are provided for co-flow and counter-flow system using genetic algorithm in MATLAB software. The optimal values of the decision variables are reported and evaluated; as a result, the efficiency of 6.56% with the power of 118.72 W can be achieved using multi-objective optimization method. In addition, the Pareto sets of optimal solutions are provided and the optimal design points are suggested for the systems. It is noticeable that, the manufacturing investment of 1.94 $/W is obtained at suggested design point for counter-flow system. It is also concluded that 37.31% decreasing in the investment and 44.81% enhancement in the second law efficiency can be achieved by reducing the power output by 9.58%. The influence of effective operating parameters on the optimal power output and second law efficiency are reported and analyzed. Considering power as objective function, the optimal length of the P-N legs are reduced by increasing the mass flow rate of the cold fluid, however other optimal decision variables are increased. Besides, increasing the mass flow rate of the cold flow to more than 0.3 Kg s can enhance the performance, significantly.

中文翻译：

---

使用遗传算法对 TEG 系统进行单目标和多目标优化以获得最佳功率、成本和第二定律效率

摘要 利用热电发电系统是将热能转化为电能的清洁装置之一，用于回收废能。在本研究中，特别关注投资和第二定律效率，考虑并优化了配备冷热换热器的 TEG 系统用于汽车尾气。首先，PN 对被离散化并划分为指定数量的元素。通过将控制方程应用于每个元件并考虑换热器的流态条件，开发了一种增强型数值算法来分析系统并计算目标函数，包括功率输出、第二定律效率、投资和复合函数这是功率、第二定律效率和投资的总和。腿的长度，PN 对的数量、换热器的水力直径和热流体质量流量被视为决策变量。使用MATLAB软件中的遗传算法为协流和逆流系统提供单目标和多目标优化。报告和评估决策变量的最佳值；因此，使用多目标优化方法可以在 118.72 W 的功率下实现 6.56% 的效率。此外，还提供了帕累托最优解集，并为系统提出了最优设计点。值得注意的是，在逆流系统的建议设计点获得了 1.94 $/W 的制造投资。还得出结论，投资减少了 37.31%，减少了 44. 通过将功率输出降低 9.58%，可以实现第二定律效率提高 81%。报告和分析了有效运行参数对最佳功率输出和第二定律效率的影响。将功率视为目标函数，通过增加冷流体的质量流量来减少 PN 腿的最佳长度，但增加了其他最佳决策变量。此外，将冷流的质量流量增加到 0.3 Kg s 以上可以显着提高性能。然而，增加了其他最佳决策变量。此外，将冷流的质量流量增加到 0.3 Kg s 以上可以显着提高性能。然而，增加了其他最佳决策变量。此外，将冷流的质量流量增加到 0.3 Kg s 以上可以显着提高性能。