Research on performance evaluation methods for silver-based electrical contact materials based on multi-dimensional parameter fusion and micro-morphology verification

Abstract To address the challenges of capturing nonlinear degradation behavior and the lack of micro-macro mapping in the evaluation of electrical contact performance for silver-based electrical contact materials, this study proposes a comprehensive assessment method integrating multidimensional parameters, state-aware weight allocation, and micro-morphology verification. Using an electrical contact performance testing system, multi-condition experiments were conducted on contacts made of different materials. Key parameters such as contact resistance, arc time, and welding force were collected in real time. Variant multiscale permutation entropy was employed to identify performance inflection points, and a dynamic weighted gray relational projection model was constructed to characterize degradation pathways. The concept of contact difference planes was introduced, integrating fractal dimension and fractal roughness features to analyze the nonlinear correlation between microstructure and performance degradation. Bayesian optimization is employed to achieve adaptive optimization of key control parameters in the failure penalty factor. Experimental results demonstrate the method’s excellent adaptability and robustness across multiple failure modes. It achieves high consistency between microstructure and macro-performance, overcoming limitations of traditional linear evaluation. This establishes a scientifically sound multi-scale assessment system, providing theoretical support for contact material performance evaluation and service life prediction.