Extensive attention has been paid to wireless power transfer (WPT) systems for electric vehicles (EV) on urban roads owing to their convenience and efficiency. To address the issues of bulky magnetic coupling structure (MCS) and fragile ferrite in WPT systems, an optimization method is proposed for multi-material magnetic cores through joint simulation using Genetic Algorithm (GA) and Finite Element Method (FEM). With ferrite volume as the constraint condition, the magnetic core structure and material proportions are optimized via discretized magnetic core coding, resulting in an iteratively obtained multi-material magnetic core layer combining ferrite and soft magnetic composites (SMCs). Compared with conventional ferrite-based MCS, the three optimized cases demonstrate ferrite volume reductions of 21.875%, 28.125%, and 37.5% respectively while preserving coupling coefficients. These findings provide valuable design guidelines for WPT system magnetic core arrangements, with potential applications extending to magnetic shielding plate optimization.