Research On Distributed Complex Event Processing In Manufacturing Internet Of Things

Industry 4.0 has brought great opportunities and challenges to the traditional manufacturing industry. Companies need to conduct in-depth manufacturing process automation, information technology transformation. Meanwhile, the market continues to face inflation, rising prices, increasing labor costs, growing user demand for personalized status quo. How to use less manpower, materials, equipment to production volume of more and better quality products to meet the individual needs has become a major problem manufacturing companies facing.Manufacturing IoT real-time collection of people, materials, equipment, and other information in the production process, such information will be processed and analyzed to achieve precise control of the production process. Manufacture IoT reduce the case of manual intervention, can quickly and effectively to ensure product quality and the safety of the device. Manufacture IoT is a good way to solve the plight of today’s manufacturing. However, the data in manufacture IoT has a mass data in the multi-source, high spatial and temporal correlation, strong uncertainties and other features. How to handle the data manufacture IoT collected, obtaining valuable information which is critical in manufacturing IoT. Existing data processing methods can not meet the strong relevance and timeliness of the data requirements for the manufacture IoT.Therefore, this paper combination of complex event processing, presents manufacturing IoT distributed complex event processing study. The main research work are:(1) Analysis of the current situation and the needs of the manufacturing data processing things, combined with complex event processing features, proposed and designed a complex event processing framework based on distributed manufacturing IoT, full use node’s computing resources and storage resources. By running a lightweight complex event processing engine in the node, acquisition data is formatted to the data model that system needs. This effectively improve the efficiency of complex event processing, while reducing the data transfer in the network.(2) For complex event detection based on event priority, how to mass flow quickly and efficiently is divided into multiple event queue is a critical problem.Design based on the priority of the bloom filter to determine engine, through the engine to quickly determine the priority of implementation events, thus improve the speed of complex event processing.(3) The design of a high throughput complex event detection algorithm. Proposed by aggregating connections to improve throughput of complex event detection. We propose to postpone the construction of pattern matching, let batch execution. By reducing the cost of traversing the connection AIS realize improve throughput.Through the above research work, can effectively resolve the manufacture IoT complex event processing systems’s problem in real-time data processing and scalability. It provides a platform and theoretical basis for further research in the manufacture IoT complex event processing.