基于信息融合与GM的石油罐区安全监控预测模型

近年来,石油罐区安全事故发生频率呈不断上升趋势。为有效增强罐区安全监控预警系统监测数据的可靠性,并实现对事故的早期预警,基于多传感器信息融合技术和灰色模型(GM)理论,建立出石油罐区安全监控预测模型。首先,研究了基于递推最小二乘法改进的最优加权融合算法,并将其作为一级(特征级)融合模型,其次,介绍分析了灰色预测理论及GM(1,1)模型的实现过程,最后建立出基于GA-BP神经网络算法的二级(决策级)数据融合模型,并得到石油罐区安全监控预测模型。

Petroleum tank farm safety monitoring forecasting model based on information fusion and GM

In recent years,oil tank farm safety accident frequency showed a trend of rising. In order to effectively enhance the terminal security monitoring early warning system for the reliability of the monitoring data,and realize the early warning of accidents. The oil tank farm safety monitoring model was built based on multi-sensor informa-tion fusion technology and the theory of grey model（ GM）. First of all,the improved optimal weighted fusion algo-rithm based on the recursive least square method was studied and it was used as a level 1（ characteristics）fusion model,secondly,the analysis of the grey prediction theory and realization process of GM（1,1）model was intro-duced,and finally a secondary based on GA-BP neural network algorithm（ decision）data fusion model was estab-lished,and the oil tank farm safety monitoring model was achieved.