Abstract
The ecological safety of plateau-characteristic agribusinesses (PCABs) encompasses not only environmental protection during production but also enhancing output without harming the ecosystem. Objectively evaluating PCABs' ecological safety levels is crucial for making management decisions and promoting local sustainable development. However, existing evaluation methods fail to incorporate expert knowledge structures, leading to limitations in both scientific validity and practical applicability. To address this gap, we propose an ecological safety evaluation method based on knowledge granularity, tailored to PCABs. The evaluation process involves the following steps: first, numerous knowledge points across macro-, meso-, and micro-levels are organized, and a hierarchical knowledge structure is designed for experts to freely select. Second, using belief distribution functions, experts provide evaluation information at varying levels of knowledge granularity. Third, expert weights and reliability parameters are determined through entropy and similarity calculations. Subsequently, the generalized combination rule is applied to progressively integrate multilevel knowledge-granularity evaluation information, from bottom to top and from individual to group, producing high-quality group evaluations. Finally, the method is applied to a typical plateau specialty agribusiness; the results indicate that its ecological safety status is "good." On that basis, we make targeted management recommendations.