Research And Implemention Of Network Traffic Predicting Algorithm Based On Smart Grid

As the direction of future energy development,smart energy has gradually become the hot spot of energy research.As an important part of the smart energy,the development of smart grid is also aroused wide attention.In the development of smart grid,A large number of intelligent devices are added to the power communication network.This leads to a sharp increase in the amount of data transmission in power communication network.How to ensure the reliability of the power communication network has become the key factor that influence the development of smart grid.Network traffic value is an important indicator of network status,and it is significative for managing communication network to predicting it's traffic value and it's increasing trend.The research of this thesis is supported by the national power grid technology project,"the research and application of the key technology of the network traffic prediction and the intelligent pipeline intelligent".The paper start with the traffic prediction technology,and it takes the integrated data network of the smart grid as the research object.In view of the characteristics of network traffic characteristics of integrated data network,a two dimensional cooperation predicting algorithm is proposed,and a software platform is developed based on the proposed algorithm which can be used for managing and predicting communication network traffic.Firstly,the paper introduces the time series predicting algorithm,then describes the application of the internet traffic in the internet,and the one step delay phenomenon in the existing network traffic prediction algorithm.Research on the traffic forecast of smart grid communication network,through the analysis of the integrated data network for smart grid data,it is found that the integrated data network traffic has obvious daily periodicity and unique trends.According to the phenomenon,this paper proposed the two dimensional cooperation predicting algorithm.Effectively solve the problem of the high prediction error of the prediction algorithm at the turning point.The simulation results show that the algorithm can effectively improve the prediction accuracy of the turning point as well as ensuring the overall prediction accuracy.Further,the paper realizes the function of real time traffic prediction based on an open source traffic management software,and the practical test shows that the two dimensional cooperation predicting algorithm has high prediction accuracy for predicting the network traffic of the smart grid communication network.