The preservation and modernization of traditional paper-cutting art face significant challenges due to its reliance on manual craftsmanship and subjective evaluation. This study proposes a computational framework integrating fuzzy analytic hierarchy process (F-AHP) and a Rime Optimization Algorithm-Back Propagation Neural Network (RIME-BPNN) to digitize and op-timize the design process for Yueqing intricate paper-cutting. First, we formalize the design op-timization problem as a multi-criteria decision-making task, where F-AHP quantitatively extracts key emotional dimensions (“Exquisite-Rough”, “Modern-Traditional”, “Elegant-Rustic”) from both artists’ expertise and consumer preferences. Second, we introduce RIME-BPNN, a metaheuristic-enhanced neural architecture that demonstrates superior prediction performance over conventional BPNN and SVR models through adaptive parameter optimization. Third, we implement a Generative Adversarial Network (GAN) that automatically generates design solu-tions by learning the mapping between F-AHP-derived parameters and visual features. Quanti-tative evaluations demonstrate the framework’s effectiveness: user studies show higher emo-tional resonance scores and greater satisfaction compared to conventional methods. The pro-posed system’s key innovations include: (1) a datadriven F-AHP method bridging subjective art evaluation with computable metrics, (2) RIME-BPNN’s superior convergence in modeling non-linear aesthetic preferences, and (3) an end-to-end pipeline from perceptual analysis to AI-generated design. This work provides a scalable computational paradigm for intangible cul-tural heritage digitization, demonstrating how hybrid AI techniques can address challenges in traditional art preservation and innovation.