基于Petri理论的水质安全预警模型构建与应用

水质安全预警对于预防水污染事故发生、保障水资源安全具有重要意义。传统的水质安全预警模型只能对渐变式水环境污染事故或已经发生且有明显征兆显现出的水环境污染事故进行预警,无法对没有明显征兆的突发式水环境污染事故进行有效预警。针对这一问题,将Petri网的跃阶思想有效地纳入传统评价指标体系中,实现对模型逻辑拓扑网络的结构性调整,并结合模糊综合评价法和BP神经网络,构建了可以对突发式水环境污染事故做出及时反应的预警模型,将该模型应用到太湖水质安全预警的研究中。结果表明:在将TN、TP、透明度等指标纳入跃阶因子的基础上,模型可对太湖多次发生的水华事件实现准确预警。该结论为研究突发水环境污染事故预警提供了新的思路和理论依据。

ESTABLISHMENT AND APPLICATION OF WATER QUALITY SAFETY EARLY WARNING MODEL BASED ON PETRI THEORY

Water quality safety early warning is of great significance for preventing water pollution accidents and ensuring water resources safety. The traditional water quality safety early warning model can only give early warning to the gradual water environmental pollution accidents or the water environmental pollution accidents that have already occurred and have obvious signs appearing, but cannot give effective early warning to the sudden water environmental pollution accidents without obvious signs. To solve this problem, the step idea of Petri net was effectively incorporated into the traditional evaluation index system, so as to realize the structural adjustment of the logical topology network of the model. Combined with the fuzzy comprehensive evaluation method and BP neural network, an early warning model which can timely reflect the sudden water environmental pollution accident was established. The model was applied to the study of water quality safety in Taihu Lake. The results showed that the model could accurately predict the occurrence of multiple bloom events in Taihu Lake, on the basis of incorporating the indexes of total nitrogen, total phosphorus and transparency into the step factors. The conclusion provided a new idea and theoretical basis for studying the early warning of water environmental pollution accidents.