Data flow adaptive clustering method in data center

With the continuous development of power grid informatization technology, the scale of data accumulated by power grid will become larger and the data dimension will be higher and higher. How to efficiently use flow clustering for power grid data mining is an urgent problem that needs to be solved. A data flow adaptive clustering method based on the power grid data middle station. The algorithm uses the high density between micro-clusters to calculate the local density peak of each micro-cluster node to adaptively generate the number of macro-clusters; Perform recursive online update; and remove the computation between edge-intersecting micro-clusters to reduce the amount of computation required to maintain macro-clusters. Theoretical analysis and experimental comparison show that the proposed method has higher performance than current mainstream stream clustering algorithms Quality clustering effect.