Multi-source Data Fusion Method in Monitoring Domain Based on Structural Similarity-Text Clustering and Ontology Glossary

Abstract

Data fusion is necessary for achieving muti-source data integration in power grid monitoring. In order to address the low accuracy of current data fusion approaches, this paper suggests a multi-source data fusion model based on structural similarity-text clustering and ontology glossary. The model uses the ontology glossary to store the clustering results and utilizes the DBSCAN algorithm based on structural similarity comparison to fuse the data. Finally, this paper compares the above model to the traditional k-means and DBSCAN model using data from the grid. The results demonstrate that the model in this paper has a higher data fusion accuracy, which suggests this model may successfully raise the level of data integration for power grid monitoring.