引入免疫机理对化工园区安全生产预警预测模型进行了研究。基于生物免疫系统与化工园区安全生产管理体系的相似性，借鉴抗原(Ag)与抗体(Ab)相互作用的机制，构建了化工园区安全生产预警指标体系，应用改进的层次分析法得到各指标权重，结果表明，影响较大的 Ag 指标有 Ag Ⅴ企(业生产危险性状况)、Ag Ⅵ企(业间相互影响状况)、Ag Ⅳ整(体规划水平)；影响较大的Ab指标有Ab Ⅰ安(全生产监管机构与人员配置)、Ab Ⅳ安(全生产监管力度)、Ab Ⅱ安(全生产规章制度)。结合Mamdani模糊模型、Elman网络算法分别建立了化工园区安全生产现状预警和未来预测模型，选取典型化工园区进行应用，获得其事故免疫力值。应用表明，该模型可对化工园区安全生产现状进行预警，对未来状态进行预测，预警预测结果与实际情况相符，并可根据模型预测结果采取相应的措施预防安全生产事故，为化工园区安全生产管理提供一种新的预警技术和方法，提升重大事故的预防预警能力。%Based on the similarity between biological immune systems and safety management systems of chemical industry parks (CIP), the interaction mechanism of antigen(Ag) and antibody(Ab) in the biological immune systems was referred to guide the development of a safety warning-and-prediction indexing system. The importance of the indexes was weighted through an improved analytic hierarchy process (IAHP). The results show that the major influence factors of Ag are AgⅤ(Risk factors of production), AgⅥ(Mutual influence) and AgⅣ(General planning), and the major influence factors of Ab are AbⅠ(Safety supervision institutions and personnel) , AbⅣ(Strengthening the safety supervision) and AbⅡ(Regulations of safety management). Inspired by the Mamdani fuzzy inference model and the Elman neural network algorithm, the safety warning-and-prediction model was established and applied in a typical CIP, and the accident immune ability was obtained. The results show that the model can fulfill the function of safety warning and prediction for the CIP, and the prediction results agree well with the actual condition. Furthermore, actions can be performed to improve the work safety management ability of the CIP following the prediction results. This model provides a new technology and methodology for safety warning of CIP, which improves the managing ability in those areas.
展开▼
