| With the rapid development of the process industry and computer application,the rapid development of enterprises and technology has brought about the optimization and upgrading of various production links.Under the background of the development of technical means and production products,the large data processing technology for various industries in the process industry came into being.How to use these data scientifically and effectively is of great value and significance to the production process and products of enterprises.The solution of processing real-time data centered on events provides new ideas for data processing.Complex event detection,as a method to detect events needed by enterprises from event flow,has been applied in many fields.Event detection for massive and complex process industry data has become a problem worth studying in recent years.As an important step in process industry monitoring equipment and production process,complex event detection can provide enterprises with the basis for fault handling and process optimization.However,due to the complex production environment of the process industry,network transmission congestion,delay,and the production process are susceptible to external interference and other reasons,Therefore,the following problems still need to be resolved:(1)The process industry aggregates and processes the event streams of different devices.The sequence of events is the key to ensuring the correctness of complex event detection.The first task is to filter and correct out-of-order events and detect intermediate complex events;(2)From simple composite events to complex events that meet user requirements,further pattern matching is needed to detect deeper levels of events;(3)Single-machine deployment is difficult to detect massive real-time events.Therefore,in order to ensure the efficient completion of complex event detection,it is necessary to adopt a reasonable distributed architecture and strategy.As an effective event stream processing technology,complex event detection conforms simple events from different event sources to specific events and rules to meet the complex events that meet the requirements.This paper focuses on the above problems faced by complex event detection in the process industry.The main work contents are summarized as follows:(1)The detection of complex events from simple events to intermediate events isaccomplished by event detection method for temporal relations.Experiments show that this method reduces the detection time and ensures further detection.(2)The event tree-based detection method is proposed,which innovatively incorporates matching results sharing and query constraints.Experiments show that the method can improve the detection rate of events.(3)The distributed real-time stream processing framework Storm and the complex event processing engine are combined to use the data distribution strategy based on event flow and rule flow,and the experimental detection results take less time.Through the research on complex event detection methods in the process industry,the process from simple events to complex events is completed,which lays a foundation for event analysis and prediction.Therefore,this paper has positive and important research significance. |
