高速公路危化品事件处置持续时间预测模型

摘要

To support emergency handling of hazardous chemicals transportation on freeways,using historical data, a precise treatment duration prediction model of hazardous chemicals traffic accidents is proposed.First,based on accident characteristics analysis,impact factors of duration are preliminarily selected.Next,correlation tests for each factor by Spearman rank correlation coefficient are carried out,and it finds out that the characteristics and leakage of hazardous chemicals and the number of needed barge trucks are typical impact factors,and they can be used as prediction model's in-puts.Then,the prediction model is developed on the basis of TSK fuzzy inference system and ANFIS method.According to the case study and its error analysis,the calculated values are in good agreement with the measured values.Besides, the reasonable amount of input parameters of this model ensures the model′s operability.Therefore,the model and its ap-plication can provide necessary references for emergency rescue of hazardous chemicals traffic accidents.%为准确预估危化品事件影响的时间、支撑高速公路危化品运输应急处置,基于危化品运输事故历史数据,构建高速公路危化品事件处置持续时间的预测模型.首先在事故特征分析的基础上,初步选取影响因子;利用Spearman秩相关系数对各因子进行相关性检验,确定危化品特性、危化品泄漏量及所需要驳货车的数量作为模型的输入变量;基于TSK型模糊推理系统,采用ANFIS方法建模;最后采用实例对模型进行验证并作误差分析.结果表明,模型预测结果与实测值吻合良好,且输入参数数目控制在合理范围内,能够为危化品事件救援提供必要的参考.