The Design Of Production Site Monitoring System Based On Esper In Ubiquitous Computing Environment

As the rapid development of modern industry, industrial products become more advanced in many different areas. The rapid development requires companies to pay attention to the production site management, which provides a reliable guarantee for improving the quality of the products. But in many traditional industrial enterprises, the lack of effective means of monitoring in the production site causes adverse consequences, such as product defects, equipment damage and so on. So the enterprises should monitor the production site to detect anomalies timely and ensure smooth production. The Internet of things realizes the integration of human society and the physical systems. It also provides a new method for real-time monitoring and management in the production site. Sensors in the production site send large amount of data in a period of time. Complex Event Processing can reveal the hidden information between the data. And the information can be used to monitor the production site real-timely.In this paper, the production of hand tools is used as an example and CEP is used to monitor the whole production process real-timely. The production of hand tools usually takes several processes and each process has its own requirements. So there is a need to use different abnormal patterns at each process. Meanwhile, based on the analysis of each process, it can be found that every process needs a period of time to prepare. If the processing standard is used to monitor during the prepare time, it will produce false positives. In order to solve these problems, this paper proposes a modeling method that the abnormal patterns can change as the production processes change.This paper presents the architecture of monitoring system and outlines the implementation process at first. It also analyzes the knowledge about CEP, and selects the Esper engine to achieve the core module monitoring system. Then this paper presents a modeling method, which divides each production process into different stages according to the characteristics of each stage. It also presents the method about judging the critical point between two stages and loading the corresponding abnormal patterns. By using this model, the monitoring of hand tools production process is more refined and reasonable. It can use the corresponding abnormal patterns according to the current stage of process, so as to solve the problem of false positives generated. In the experimental analysis section, it designs the core module in detail and prepares the experimental data. Finally, it presents the test results of core module, including the accuracy and efficiency of different types of abnormal patterns. Test results show that the proposed method can avoid false positives effectively and improve the accuracy of monitoring.