Electromagnetic parameter prediction and multi-band absorption design of ZnO@YCO composite materials based on an SSA-GRNN-PSO data-driven framework

As electromagnetic environments become more complex, the design of novel electromagnetic wave absorbers (EWMs) that combine low-frequency absorption with multi-frequency response remains a significant challenge. Experimental validation demonstrates that controlling the crystallinity and morphology of the ZnO coating layer through process temperature can significantly influence the electromagnetic properties of YCo composite materials. Nevertheless, the establishment and description of the intricate relationship between process temperature and electromagnetic response remains a formidable challenge. In order to address this issue, a data-driven modeling and inverse design framework (SSA-GRNN-PSO) integrating sparrow search algorithm (SSA), generalized regression neural network (GRNN), and particle swarm optimization (PSO) is proposed based on limited experimental data. This framework facilitates precise prediction of the complex electromagnetic parameters of ZnO@YCo composites at varying processing temperatures. When combined with a multilayer structural design concept, it achieves multi-band inverse design. The final composite material, with a thickness of 19.1 mm, exhibited strong multi-band absorption characteristics, with a bandwidth of 7.452 GHz, and a reflection loss of − 25.47 dB in the 2–18 GHz range. Moreover, an optimized design with a thickness of 5.4 mm was achieved, which features a bandwidth of 7.416 GHz and a reflection loss of − 15.72 dB.