Construction of Unified Data Model for Smart Grid Considering Frequent Item Sets
In the smart grid data modelling,there are many problems such as complicated model types and long calculation time.A unified smart grid data model considering frequent item sets is constructed.By considering the frequent item sets,an im-proved Aprioro algorithm is designed.Four aspects are considered in the improvement,mapping the transaction database through the matrix idea,optimizing the connection step,optimizing the pruning step compressing transaction database and im-plementing file data mining in smart grid online analysis system.For the data content of the mined file,a series of preprocess-ing is implemented,including time-series data repair and time-series data screening.The unified data model of smart grid is built through distributed memory data grid technology.In the model design,the serialization method is writeData method,and the deserialization method is readData method.The test result is that the model analysis and calculation time is 0.35 s.When the number of buses updating data reaches 70,the model can maintain the updating frequency about 60 000 times/min.when the number of nodes updating data reaches 800,the model can maintain the updating frequency more than 110 000 times/min.The future state power flow calculation and sensitivity calculation of the model are both low in time-consuming.
frequent item setssmart gridimproved Aprioro algorithmdistributed memory data grid technologyunified data model
万方数据
维普
