Research On Power System Architecture And Dispatching Method In Cloud Computing Environment

With the expansion of the scale of the power system and the improvement of the power grid,the data collected in the power grid is getting bigger and bigger,and the power system gradually enters the era of large data.The existing scheduling architecture can not satisfy the fastness of the large data processing.Because cloud computing technology has the characteristics of fast processing large data,it can be used to solve the problem of large data processing in power system.This paper mainly studies the power dispatching algorithm based on cloud computing technology.This paper first introduces the background of the topic,and the domestic and international research status in the grid scheduling architecture and cloud computing scheduling algorithm two aspects.On this basis,the power grid dispatching structure under the cloud computing environment is designed from the aspects of power system scheduling software and hardware.In hardware,from the four scheduling subsystem to start,cloud computing integrated the existing power grid scheduling resources to connect the scheduling system cloud computing resources.In software,establishing the cloud operating system to manage all the resources based on the integrated hardware,constructed a power system advanced application software cloud computing operating environment.Aiming at the establishment of the communication system under the grid computing architecture of the cloud computing,the processing tasks are allocated based on the greedy algorithm,and the dynamic resource scheduling algorithm is designed on the basis of greedy algorithm.Through monitoring of server performance changes and data upload/download speed changes in different regions,the use of secondary smoothing index model predicts the resources,and performs real-time assignment tasks based on the predicted results.The simulation results show that the proposed algorithm has better allocation strategy than FIFO algorithm and static greedy algorithm.Aiming at the large-scale power system,this paper studies the parallel algorithm of static security analysis,performs parallelization based on the mutual independence of different grid states in the expected accident concentration,and simulates it in the cloud computing environment.The results show that matpower2869 node calculating time under the dynamic resource scheduling algorithm is about 60%less than the FIFO algorithm,about 80%less than the stand-alone mode.