Power Incomplete Data Clustering Based on Fuzzy Fusion Algorithm

With the rapid development of economy, scale power grid is expanding. The number equipment that constitutes has been very large, which makes state data grow explosively. These multi-source heterogeneous have differences, lead to variation in process transmission and preservation, thus forming bad information incomplete data. Therefore, research on integrity become an urgent task. This paper based characteristics random chance Spatio-temporal difference system. According sources massive generated by equipment, fuzzy mining model established, divided into numerical non-numerical Take text defects as material. Then, Apriori algorithm array used mine deeply. strong association rules are obtained analyzed. From change trend NRMSE metrics classification accuracy, most filling methods combined with two frameworks this method usually show a relatively stable trend, will not fluctuate greatly growth missing rate. experimental results proposed can effectively improve effect existing sets, fluctuates increase rate, is, improvement for higher than 4.3%. Through clustering technology studied paper, more innovative assessment smart reliability operation carried out, good value reference significance.