Kalkan, Orhan; Demir, Fatih

Lithium-ion batteries are frequently preferred for electric vehicles and diverse applications due to their high energy density, high specific energy, and superior efficiency. Despite these advantages, they generate substantial heat during charge and discharge at high C-rates, which can lead to thermal runaway. Thus, an appropriate Battery Thermal Management System is necessary to prolong their lifespan and maintain performance and capacity. This study considers a passive cooling technique for cooling a Lithium-ion battery using fins. The effect of fin geometric parameters on battery thermal management was numerically investigated, and optimal values were determined through multiple optimizations, thereby contributing to the literature. For this purpose, analyses were obtained for varying diameters (20 mm-36 mm), thicknesses (0.5 mm-2 mm), and numbers (3-9) of cylindrical fins placed on a cylindrical battery at a high discharge rate of 5C. In the results, battery surface temperatures, battery temperature difference, and mass were evaluated in detail. Next, the maximum and average battery temperatures, the maximum and minimum temperature differences for homogeneity, and mass were selected as objective functions for multiple optimizations. The I-optimal algorithm based on the Response Surface Method was used to create a Design of Experiment. Optimization parameters and levels were determined by numerical analysis. The mathematical models derived for the selected performance parameters (maximum temperature (Tmax), average temperature (Tavg), maximum and minimum temperature difference (∆T) for homogeneity, and mass were tested using ANOVA, with prediction accuracies of 0.9923%, 0.9866%, 0.9381%, and 0.9314%, respectively. The results indicate that increasing the fin diameter (from 20 to 36 mm) decreases Tmax by 3.4% and Tavg by 3.7%, but increases mass by 12.2% and ∆T by 5 K to 11 K. Moreover, as fin thickness increases from 0.5 mm to 2 mm, Tmax and Tavg decrease by 2.2%, but increases mass by 21% and ∆T by 9.8 K to 10.8 K. Furthermore, as the number of fins increases from 3 to 9, Tmax and ∆T decrease by 6.0% and 36.3%. Increasing the number of fins is more effective at increasing homogeneity than increasing their diameter and thickness. Optimization results show that with a fin diameter of 32.4 mm, a thickness of 1.35 mm, and 9 fins, the battery temperature can be maintained below 349 K under these optimal conditions.

Yuksel, Mehmet Erkan; Sevgen, Selcuk; Yıldız, Eyyüp

Loss functions are crucial in training generative adversarial networks (GANs) and shaping the resulting outputs. These functions, specifically designed for GANs, optimize generator and discriminator networks together but in opposite directions. GAN models, which typically handle large datasets, have been successful in the field of deep learning. However, exploring the factors that influence the success of GAN models developed for limited data problems is an important area of research. In this study, we conducted a comprehensive investigation into the loss functions commonly used in GAN literature, such as binary cross entropy (BCE), Wasserstein generative adversarial network (WGAN), least squares generative adversarial network (LSGAN), and hinge loss. Our research focused on examining the impact of these loss functions on improving output quality and ensuring training convergence in single-image GANs. Specifically, we evaluated the performance of a single-image GAN model, SinGAN, using these loss functions in terms of image quality and diversity. Our experimental results demonstrated that loss functions successfully produce high-quality, diverse images from a single training image. Additionally, we found that the WGAN-GP and LSGAN-GP loss functions are more effective for single-image GAN models.