With the popularization and development of computer technology and information technology, the amount of information data is drastic increased. The processing and management of information has also been more and more important. Data mining as the main technical of information processing and knowledge discovery has also been continually developed. In this paper, use the data mining technology like T-S fuzzy neural network and rough set theory, a non-linear function and the data of boiler combustion system are data mining objects, do research of non-linear system modeling, boiler combustion system modeling, and the neural network's adaptability about noise and attribute reduction algorithm. Find out a method that with a better adaptation about nonlinear with noise system modeling, and optimization and simplification the complexity of neural network modeling in the latter part by the attribute reduction algorithm, and improved the accuracy and efficiency of data mining. |