Research And Application Of Time Series Association Rules In Safety Warning Of Oil Construction Operations

Safety production is the premise of the sustainable development in the petroleum industry. The oil construction work has many industries features, like field work site, environmental complexity, mobility, work long hours, many types of work, many type of process etc, so it is easy to cause an accident. Early warning system can make a whole judgment to the Status of implementation and historical trends, according to the forecast make accidents inhibit in the bud.Specific research work is as follows:(1) Study the potential accidents in oil operation, according to the features of oil operation and the complex relationship of attributes, we pre-process the original database and reduce the attribute in temporal database.(2) After the attributes reduction, combined with time series and association rules, mining the temporal association rules of single attribute and multi-attribute, forecast the future development trend of attribute.(3) Analyzed the temporal association rules then removed the conflict rules.(4) Based on the petroleum construction safety system, designed the safety warning module.This paper combined with the project of "production safety early warning system", studied and analyzed the historical data, extracted the temporal association rules from temporal database. Then predict the future trend of attribute index. It aims at providing reference data to policy makers, reducing the occurrence of safety accidents.