Research On Event Fusion Method For Industrial Big Data

Under the background of the rapid development of information technology and industrial automation,industrial enterprises and industrial technology developed rapidly,and all aspects of industrial processes were optimized and upgraded.At the same time,the rapid development of products and technology,all kinds of big data came into being in the field of industry.And how to use these industrial data scientifically and effectively is of great significance to the development of enterprises,product development,process optimization and so on.The data contained in the industrial big data is of various data types,complex structure and a wide range of sources.At the same time,the accumulation of data makes the processing cycle longer.The structure is complex,so that the processing difficulty and the amount of calculation is increased.The processing of real-time data rather than storage data also increases the difficulty of calculation and processing.Therefore,it is of great theoretical value and practical significance to optimize the industrial process scientifically and reasonably.In view of the current situation that the effect of monitoring and processing real-time data needs to be improved,this paper proposes an event detection method based on directed graph by using the complex event processing technology.Aiming at the present situation that the data can not be effectively connected,this paper presents an event correlation method based on time series.The main work and contributions of this paper can be summarized as follows: Firstly,introduce the related background of industrial big data,analyze the value of industrial big data combined with the status at home and abroad.In the current industrial big data and its event concept is vague and abstract to define its concept,and detailed description of event types,relationships and so on.Secondly,through the application of complex event processing technology,describes the event detection method based on directed graph,explains the model of directed graph and procedure of testing one by one.Thirdly,builds the model on the basis of the introduction of event relation,correlation processing by using the relation of time series,and then to make event correlation analysis.Finally,by comparing Hadoop and Spark platform,selects the Spark platform to process experimental data,and then to analyze and summarize the experimental results.Experiments show that the proposed processing method based on directed graph and based on time series can meet the demand of event fusion for processing massive industrial data,and optimizeindustrial process under the premise of ensuring normal production.