Research On The Application Design And Data Processing For Industrial Real-Time Database

With industrial process systems increasingly becoming more and more complicated, the industrial real-time database has been widely used for industrial process control and monitoring to guarantee that the systems are operating safely, stably and with high efficiency. Furthermore, such real-time database is gradually becoming an indispensable basic data platform for advanced control and process monitoring as well as plant-wide information management, which results in significant economic and social benefits. The industrial real-time database system is provided with unique time characteristic and system structure, powerful functions for real-time data collection and historical data preservation, and abundant Application Programming Interface and Data Management modules, hence, based on a thorough comprehension of its various functional modules, it is of great application value to carry out application design and development for the database and further effectively make use of the data collected or preserved therein for process control and monitoring. With the development of a polyethylene production control and monitoring system for a petrochemical plant as the background, this thesis focuses on the application design and data processing for industrial real-time database.The main contents and results of this thesis are as follows:(1) Basic concepts of real-time database are introduced, such as its generation, functions, development trend, hierarchical and management structure. Depended on a detailed description of the characteristics of real-time database, a comparison between two typical real-time database systems: PI and InfoPlus.21 are presented. The current application of real-time database in industrial process control is also reviewed.(2) The basic steps of application design of real-time database in industrial control are presented as well as a detailed description of the database design, which are illustrated by a practical application of PI database design in a project development(3) Both the OPC interface of real-time database and DCS system and the system configuration are described in detail. Then the API and SDK methods for interface design of real-time database and application are introduced. A software for managing the PI system by means of SDK method is developed, by which querying and editing the points and data in the database are easy to be realized. The efficiency and application environment of the two methods are valued in the end.(4) The difficulties of effectively using data in real-time database for modeling are presented, e.g. there generally exist missing data, redundant data, or abnormal data in database and the data dimensionality is always very high. The general steps and methods such as linear and nonlinear interpolation, data filtering, principal component analysis are adopted to overcome the foregoing problems, and a software for data preprocessing integrating the described steps and methods is developed under a combined development environment with the advantages of Excel, VB and Matlab. Its validity and adaptability are verified through the application to real projects.(5) Based on the application design, interface design and the data preprocessing, a polyethylene production monitoring platform is developed, which is running safely in industrial field. According to author's own experience, the further development of the system is forecasted.