Research On Application Of Cloud Computing In Power System

The rapid development of modern grid brings a lot of convenience. But at the same time, some problems occur that the existing power system network cannot meet the demand of the modern power grid construction. With the continuous development of the grid, the amount of equipment condition monitoring data has increased so much that far beyond the processing capabilities of traditional grid. The conventional data storage and processing methods cannot meet the demand. In the face of large scale power system data generation, collecting, storing and analyzing the large scale power system data accurately and timely are always the goal that the power system construction pursues.Power system monitoring data is continuously generated in the form of data streams. It belongs to the big data which has the characteristics of variety and large amount. This puts forward higher requirements to power system data management system. In the traditional way, mutual backup front-end processors acquire the monitoring data and send it to the relational database. Because of poor system scalability and concurrency, the relational database is difficult to meet the needs of the monitoring system for large data storage and management. The data analysis and calculation is mainly using high performance computers for serial computing, or using communication middleware to build a parallel computing platform for parallel computing. But there exists some problems to deal with massive data, such as low efficiency, time-consuming to build the platform, inadequate risk control and so on.In order to solve the above problems and achieve efficient processing of monitoring data, this paper firstly analyzed the characteristics of the existing data collection, storage and analysis solutions. Then took into account the company’s management of all levels of power grid companies, designed a solution based on hadoop cloud computing technology for Power System monitoring data parallel collection, distributed storage and parallel computing.Secondly, based on the proposed solution, analyzed the collection, storage and computing for grid monitoring data in detail, realized the distributed data storage based on HBase and data processing based on MapReduce.Finally, simulation experiments were carried out for the proposed solution. In the laboratory environment, built a small Hadoop cluster platform which includes multiple machines. Through simulation experiments, this paper compared the speed of data collection, storage, access, analysis and calculation of the proposed solution with the traditional ones. The results show that the greater the amount of data, the better the performance of parallel access, analysis and computing. The results also verify the feasibility and efficiency of the proposed solution.