Optimizing Hybrid Active–Passive Thermal Management of Prismatic Li-Ion Batteries Using Phase Change Materials and Porous-Filled Mini-Channels

Abstract

Tackling climate change is crucial, and electrifying the vehicular transportation sector is essential to reduce greenhouse gas emissions. Lithium-ion (Li-ion) batteries are highly efficient for electric vehicles (EVs) but face challenges such as thermal management, risk of thermal runaway, and high costs of lithium and cobalt. Overcoming these challenges is vital for the widespread adoption of hybrid and EVs. To overcome this drawback, this article proposed a large-kernel attention graph convolutional network (LKAGCN) with leaf in wind optimization algorithm (LWOA) named as LKAGCN-LWOA technique, which enhances the thermal management of prismatic Li-ion batteries by integrating both active and passive cooling techniques. The system incorporates phase change materials (PCMs) with porous-filled mini-channels to regulate battery temperature effectively. The LKAGCN analyze thermal properties, battery conditions, and PCM characteristics to predict and optimize the thermal behavior of the battery pack using LWOA. The proposed methods tune the parameters of the hybrid thermal management system, ensuring efficient thermal regulation and improved performance. The proposed method is compared to various existing methods such as convolutional neural network (CNN), Taguchi method, and Finite element model (FEM).