Plant Operation Working Condition of the Optimal Combination of External Research Division

Abstract

On the basis of problems which is based on external power station power plant operating conditions partitioning historical operational data analysis, it points out the key of data mining results will depend on the adaptability of the data mining algorithms. To solve the above problems, the new k-mean algorithms and equal width method which are suitable for power plant historical operating data divided external conditions are proposed. First of all, two methods are study. One is calculation method combines statistical theory of k-means clustering algorithm for the initial number of cluster centers. And another is the equal width method determination of the number of clusters and the interval size. Secondly, new K-means algorithm is to apply the history running data of the power stations in order to mine the unit load, the external condition of coal quality characteristics. Meaning while, the way to mine the external environmental temperature condition is the proposed method of equal width. Finally, to provide data for the reference site operating personnel, by combining obtain the optimal combination of external operating conditions described unit operation.