Research And Applicationg Of Big Data Analysis On Status Monitoring System In Smart Substation

With the rapid development of the related areas technology of cloud computing, big data processing and condition monitoring technology, they are widely used in smart substation status monitoring. Substation equipment status monitoring data shows exponential growth, it brought great challenges in storage, management and analysis capabilities for traditional capabilities substation equipment condition monitoring data platform. and computing and storage capacity of the existing hardware provided has been difficult meet the requirements of future power enterprise data storage optimization and parallel processing. Used the big data analytics technology into substation equipment condition monitoring, can improve the utilization of data, information sharing and data analysis capabilities of condition monitoring.Firstly, combining with the demand of business development for electricity a traditional scheme of substation equipment condition monitoring platform was given. Then given the poor storage performance, low efficiency analysis for traditional state monitoring platform, a substation equipment condition monitoring data platform in big data environment, three kinds of distributed data analysis solutions. Such as the Hive of relational on-line analytical(ROLAP), the Impala of relational on-line analytical(ROLAP), the HBase of multi-dimensional online analytical(MOLAP) were proposed. At the same time a distributed ROLAP star model and the process of state monitoring platform were designed. Conventional data model involves a lot of operations between dimension table and fact table mapping, leading to slow and expensive data analysis response speed. Aiming at this problem, this paper proposes a connectionless hierarchical coding substation equipment condition monitoring data model, the model coded the hierarchical information of the dimension table,compression to the fact table, reduce a large number of connections, and to optimize its performance.Because the Hadoop platform has characteristic of easy extension, high reliability and high efficiency equipment utilization. This paper uses Hadoop distributed cluster to construct substation equipment condition monitoring experiment platform, uses Hive and Impala large data analysis tools, uses different scale of monitoring data sets to test the monitoring data loading, roll-up and storage cost. The experimental results show that the data load slower than conventional model, but the drill-up performance and storage costs are superior to conventional analog model. The effectiveness and feasibility of the model is verified by experiment, this model can solve lots of tedious list connection operation between the distributed ROLAP,which is more suitable for large-scale, distributed cluster to perform data analysis, brings a new train of thought to the substation equipment condition monitoring large data analysis.