The Application And Research Of Data Mining For Intelligent Monitoring And Managing System

During modern industry production time, industry production takes on a shape of centralized and mass production. This has a firm relationship with the rapid application of science and technology result to industry product. A perfect fitness is the computer technology introduced to industry technology. Maturity of industry control and digit monitor systems not only increases the efficiency of production and quality of the products, but also reduces the intensity of the workers. Meanwhile it brings a higher requirement to the safety of production.The industrial intelligent system of monitoring and managing is the efficient scheme to resolve this issue. The system can not only evaluate the production form the collectivity point, but also provide potent suggestions. The key to construct the whole system is to collect the impactful control measures. The analysis of the collecting data in the existing system can achieve this goal. This paper has combined the feature of the Active Lime Rotate Kiln, refined the abundant of useful information by using various data mining algorithms and tools. A data mining engine is proposed on the basis of integrating all the tools.This paper has researched the data mining technology, represented the type of tasks, processes and steps of data mining. Especially, the components of mining algorithms have being described in detail. The regular analysis of statistics, analysis of clustering, association rule, has been mentioned as well. Then raised the design of industry intelligent monitoring and managing system structure, also analyzed the position of data mining engine in the system and the roles in different phases. Based on the research of the previous data mining technology and the system structure design, the mining of the historical date has been started. The researching procedure of each phase is described amply during the process of mining. In the first phase, the date in the existing system and the principle of production line should be analyzed, and the expert oriented method has been adapted to obtain knowledge. Then, some models are built via mathematical methods. This method can conclude good result of analyzing and models of control, when the factor is one to one. However, when the condition changed, the demerit of mathematical method is exposing. Due to the limitation of mathematical method, other methods should be applied to the system, and the finding pattern data mining algorithm is a good choice. This is the kernelcontent of this paper. During the process of finding pattern, the efficient information extraction strategy has been applied;clustering analysis is used to define the security range of each monitoring point under different output. A more efficient algorithm is proposed, according to the demerit of the existing algorithm to identify the security rules and perfect the models. On the basis of writer's researches about various data mining methods and tools introduced through out the paper, a model aiming at data mining is raised. The model with excellent expansibility orients problems more effectively, solving all kinds of problems of date mining.