Graphene aerogel, known for its excellent dielectric properties and porous structure, demonstrates outstanding electromagnetic wave absorption. In this study, we fabricated lightweight (11.16 mg/cm³) graphene composite aerogels incorporating reduced graphene oxide (RGO), manganese ferrite (MnFe₂O₄), and polyimide (PI) with oriented pore structures. The addition of MnFe₂O₄ enhances impedance matching and synergistically combines dielectric and magnetic loss mechanisms, significantly improving absorption performance. The composite exhibits an effective absorption bandwidth of 9.216 GHz, covering nearly the entire X and Ku bands (2–18 GHz), with a minimum reflection loss of −75.643 dB. In the 18–26.5 GHz range, the effective bandwidth reaches 7.80 GHz, with a minimum peak of −58.862 dB. At a thickness of 2–2.4 mm, the bandwidth extends to 8.5 GHz, covering the full 18–26.5 GHz range. With a thickness of 1.46–1.8 mm, the bandwidth reaches 13.5 GHz, spanning 26.5–40 GHz, and a minimum peak of −79.560 dB with an 11.30 GHz bandwidth. The oriented porous structure enhances electromagnetic loss, as confirmed through experimental and theoretical analysis. The composite also exhibits excellent mechanical strength and thermal insulation, attributed to its oriented microchannels. This work provides valuable insights for designing lightweight, broadband, and efficient graphene-based absorbers.

Textile memristor is the decision-making center of a bionic neuromorphic intelligent textile system and presents a promising opportunity for the incorporation of intricate information processing components within the architecture of flexible electronic textiles. This architecture, characterized by its cross-bar configuration and adaptability, positions textile memristors as a highly promising avenue for the future advancement of wearable electronic devices. Here, carbon fiber is chosen as electrode, and the memristor functional layer is made up of a high-performance and reliable BiOI/TiO₂ heterogeneous interface synthesized hydrothermally. The textile memristor exhibits an average set voltage of ≈ 0.78 V, set voltage deviation of 12.3%, long retention time (> 10⁴ s), and high on/off ratio (> 10⁵). Simultaneously, the observed self-rectification, room-temperature negative differential resistance (NDR) effect, and biological synaptic phenomenon offer the possibility of further neuromorphic applications. Subsequently, the pressure sensor was integrated onto the textile memristor to construct an intelligent textile system. Especially through self built datasets and neural network calculations, the gesture recognition rate of the system reached 98.5%, maintaining a high level under the influence of 20% Gaussian noise (64%). This textile memristor arrays supply great memristive performance, steady device distribution, and good mechanical strength, indicating a reliable direction for the next generation of integrated textile electrical systems.