Research On Fast Process Of Alarm Data Of Electrical Equipment Based On Storm Platform

With the rapid growth of data scale in electric power industry, Electricity Big Data era has arrived. The construction of Smart Grid requires higher performance of real-time processing. In recent years, cloud computing technology develops rapidly and distributed processing platforms continue to emerge, which offers effective infrastructure for the real-time process demands of massive power data. Storm is a distributed real-time computing framework. It has achieved great success in internet field since it opened source in 2011 for its characteristics of real-time processing capability, high reliability and scalability. Thus we consider introducing Storm to the real-time analysis and processing of grid monitoring data.In this paper, the research about large-scale grid data process is reviewed and the problem of real-time process of massive monitoring data is discussed in particular. Then current big data processing platforms are summarized. Combine with the application requirements of quick response of large-scale alarm data, Storm distributed platform is chosen to carry on the research.The key points of the research are Storm clustered environment deployment and the immigration of existing business to Storm platform. Firstly, the Approximate Entropy algorithm was designed based on Storm for fast feature extraction. The algorithm was divided into three main steps: data collection, mathematical analysis and result output. All procedures were realized under the pre-defined component and submitted to the Storm cluster for execution. Then the DBSCAN clustering algorithm was designed on Storm for clustering analysis of grid monitoring data flow. The classification results are processed efficiently and the abnormal data are aggregated for alarm. Besides, the works carried on the platform include data limit check and data filtering. Through comparative analysis and performance monitoring, the system shows the advantages of low latency and high throughput capacity.The research offers a new solution for the real-time process of electrical equipment monitoring data under the background of Electricity Big Data. It shows the possibility of applying Storm to the real-time processing business in Smart Grid and Storm has broad application prospects.