摘要
由于单一传感器在石油罐区安全监控中容易受到外界因素影响从而产生误差,为提高传感器检测结果的可靠性和罐区安全监控预警的准确性,基于多源数据融合技术,建立罐区安全状态预警模型。首先,介绍了多源数据融合技术的3个层级:数据级融合,特征级融合和决策级融合,以及目前各领域常见的数据融合方法;其次,建立了基于最优加权融合算法的一级融合模型和基于BP神经网络算法的二级融合模型;最后,得到石油罐区安全监控数据融合模型,并为进一步的实践应用打下了理论基础。
Due to the single sensor in the safety monitoring of oil tank farm is easily influenced by external factors and resultes in errors, in order to improve the reliability of sensor detection and the accuracy of tank farm safety mo- nitoring, based on the multi-source data fusion technology, an early - worning model of safety status in tank farm was established. Firstly the 3 levels of multi-source data fusion technology were introduced including data level fu- sion, feature level fusion and decision level fusion, as well as the common data fusion methods. Secondly the 1st level fusion model based on optimal weighted fusion algorithm and 2nd level fusion model based on BP neural net- work algorithm were established. Ffinally safety monitoring data fusion model of oil tank farm was obtained, which provides the theory basis for further practice application.
出处
《中国安全生产科学技术》
CAS
CSCD
2014年第3期90-94,共5页
Journal of Safety Science and Technology
基金
"十二五"国家科技支撑计划项目(2012BAK03B03)
