Modeling And Optimization On The Hot-Side Of Automotive Exhaust-Based TEG Based On Surrogate Model Method

According to some statistics, on the deadline of 2014, the ownership of traditional internal combustion engine vehicles in China is 154 million. Moreover, on the background of promoting energy saving and emission reduction by our government, the technology of automotive exhaust-based TEG is a most enjoyed potential development direction. The TEG, which turns the thermal energy of exhaust gas into electric energy, will enjoy broad market prospect and good economic returns. Therefore, it should be given a close attention and in-depth research. Furthermore, one of the core research directions is to study the hot-side of TEG, in order to find out the way to develop energy conversion efficiency of TEG.For the traditional optimization design relying on design experience and bench test too much, the Surrogate Model Method is innovatively used to optimize the structure and performance of heat-exchanger in this paper. In this way, it can overcome the limitation of traditional optimization design and promote the optimization to be more scientific and efficient.In this paper, the 4 Surrogate Models(Objective Functions), which can precisely reflect the performance of the heat exchanger, are innovatively defined. The 5 Independent Variables(guiding fins' length, width, height, distance and angel) are scientifically defined and are used to conduct Orthogonal Experiment. The Response Surface Model is used to fit the Approximate Functions and their fitting accuracy is verified by Latin Square Experiment. The AMGA Algorithm is used to do multi-objective optimization, in order to gain the optimal structural parameter of heat exchanger. The association graph is used to find out the relationship of each Surrogate Model. The 6-Sigma Quality Method is used to develop the robustness and confidence coefficient of the Surrogate Model. The simulation calculation and bench test are scrupulously conducted to verify the goodness of fitting coefficient among optimal results, simulation results and bench test results.The results of this research show that by using Surrogate Model Method, the multi-objective functions can be obtained, which can accurately evaluate the performance of the heat exchanger. And its fitting-coefficient is very high. By using AMGA Algorithm to optimize the heat exchanger, its optimization result is ideal and credible, and the simulation result is very close to bench test result. Applying 6-Sigma Quality Method for quality analysis, the optimal model has high robustness and excellent confidence coefficient.In this paper, the innovations are as follows: Firstly, the Response Surface Model is used to get the Surrogate Model, which can decisively reflect the thermal performance of heat exchanger; Secondly, the AMGA Algorithm is used to do multi-objective optimization of the heat exchanger's thermal performance; Thirdly, the 6-Sigma Quality Method and bench test are conducted to verify the goodness of fitting coefficient and robustness of optimization results.In this paper, aiming at the optimization design of the heat exchanger in automotive exhaust-based TEG, the Surrogate Model optimization method theory is used. It can greatly enhance the confidence and avoid a lot of simulation and bench test. Therefore, it will enjoy a good application prospect.