基于危害物风险综合评价的粮食抽检决策研究

Research on grain sampling decision-making based on comprehensive risk assessment of hazards

中文关键词:  粮食危害物  风险综合评价  膳食暴露评估  贝叶斯网络  抽检决策

英文关键词:Food hazards  Comprehensive risk assessment  Dietary exposure assessment  Dietary exposure assessment  Bayesian network  Sampling decision-making

中文摘要:

      目的 解决传统粮食危害物风险评价不够全面，粮食抽检决策随机性和主观性强的问题。方法 提出了一种基于危害物风险综合评价的粮食抽检决策方法。首先，构建多维梯形云模型实现了危害物的可能性风险评价；同时构建膳食暴露评估模型，实现了危害物的危害性风险评价，再利用风险矩阵将其与危害物可能性风险评价结果相结合，实现粮食危害物的综合风险评价。其次，在决策支持体系的基础上，明确提供基于季度、地域和危害物抽检排序的决策支持。然后进行贝叶斯网络结构学习与参数学习，并根据粮食危害物的综合评价结果进行贝叶斯网络推理，提供精细化抽检决策支持。结果 通过对2018—2019年中国稻米危害物抽检数据的分析，这种综合风险评价方法能够有效的评估稻米危害物的风险水平。同时基于贝叶斯网络，能够得到精细化的抽检方案。 结论 本研究提出的基于危害物风险综合评价的粮食抽检决策方法可以为食品安全监管部门的抽检工作提供准确、高效的决策依据，一定程度上优化抽检资源的分配，使得抽检更加科学与高效。

英文摘要:

      Objective To address the limitations of traditional grain hazard risk assessments, which are often insufficiently comprehensive, and the strong randomness and subjectivity in grain sampling decisions. Methods This study proposes a grain sampling decision-making method based on comprehensive hazard risk evaluation. First, a multidimensional trapezoidal cloud model was constructed to assess the likelihood of hazards, while a dietary exposure assessment model was developed to evaluate the severity of hazards. By combining the results of these two assessments using a risk matrix, a comprehensive risk assessment of grain hazards was achieved. Second, a decision support system was established to provide guidance based on the ranking of sampling priorities by quarter, region, and hazard type. Bayesian network structure learning and parameter learning were then carried out, and Bayesian network inference was performed based on the comprehensive hazard risk evaluation results to offer detailed sampling decision support. Results An analysis of China’s 2018-2019 rice hazard sampling data demonstrated that this comprehensive risk evaluation method effectively assesses the risk level of rice hazards. Additionally, using Bayesian networks, a refined sampling plan can be generated. Conclusion The grain sampling decision-making method based on comprehensive hazard risk evaluation proposed in this study provides accurate and efficient decision support for food safety regulatory authorities. It optimizes the allocation of sampling resources, making the sampling process more scientific and efficient.