A smart sensor-based sensing fabric surface texture identification method

By constructing a signal enhancement mechanism based on the micro-contact topology and symplectic geometric phase space of the interface, combined with an improved DeepWalk model and formal concept analysis, the problem of structural missing feature representation in texture recognition is solved, thereby improving the robustness and accuracy of texture recognition.

CN122336412APending Publication Date: 2026-07-03HUBEI UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-08
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing texture recognition methods suffer from structural loss of feature representation when processing low signal-to-noise ratio signals. They neglect the complex topological structure of the contact interface of the sensing fabric and the microscopic geometric characteristics of the triboelectric signal, resulting in insufficient robustness and reliability of texture recognition.

Method used

By constructing a signal enhancement mechanism for the microscopic contact topology and symplectic geometric phase space of the interface, and using an improved DeepWalk model and formal concept analysis logic reasoning, combined with tensor ring decomposition and tangent space alignment, the interlocking degree and high-frequency component features are extracted, the high-frequency components of texture friction are reconstructed and adaptive gain control is performed, thus realizing a closed loop from microscopic physical contact modeling to deep geometric topology description.

Benefits of technology

It significantly improves the robustness and accuracy of texture recognition, realizes deep decoupling and structured characterization of triboelectric signals, and ensures the consistency of recognition results with the surface under test.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122336412A_ABST
    Figure CN122336412A_ABST
Patent Text Reader

Abstract

The application discloses a kind of sensing fabric surface texture identification methods based on intelligent sensor, it is related to texture identification technical field, comprising the following steps: S1, output triboelectric signal sequence;S2, output texture friction high-frequency signal sequence;S3, construct standardization microtexture feature vector;S4, utilize improved DeepWalk model to combine anisotropic diffusion kernel simulation manifold diffusion process, by tensor ring decomposition and tangent space alignment aggregation mapping is low-dimensional embedding, based on tangent space relative geometric distance executes classification decision, output initial texture identification result;S5, output logic optimal texture category;S6, output final texture identification result;S7, the consistency of identification result and the surface to be measured is verified by tactile interaction.The application overcomes the limitation that prior art ignores contact mechanism, lacks essential description and lacks verification means, significantly improves the robustness and accuracy of texture identification.
Need to check novelty before this filing date? Find Prior Art