Method And Implementation Of Performance Prediction For Smart Power Distribution Communication Network

Power Grid is an important support for the developing of a country, Electric power communication network provides a reliable technical support for the safety production and monitoring of power grid. With the Implementation of the strategy of smart grid, power grid becomes more informational, more automatic and more interactive, and also the requirement of interaction during Smart Grid power distribution improves gradually. As an important component of Smart Grid, smart power distribution communication network is a significant condition for building Smart Grid. Smart power distribution communication network system provides the methods and tools for improving the modernization of Smart Power Distribution communication network. It has important significance about providing safety production, integrating resources, increasing quality of service and improving the process of intelligentizing power distribution.As a component of network management system, performance management provides the index and running report of network to the operation and maintenance personnel, and provides the reference data for network operating, managing and planning. As an important function of performance management, performance statistical analysis flinction analyzes the performance data from history and current sides, such as real time performance management and historical performance analysis. But it doesn’t use the performance data for digging in deeply, this causes the waste of performance data. Hence, fully utilizing the performance database, using analysis method, from future side to find the relative knowledge of these data, not only make it possible to fully utilize performance data, but also improve the early warning ability of power distribution communication network management system.In this paper, first we introduce the Smart Grid communication network and discuss the communication technologies in detail. We also propose a layered architecture of synthetical network management system. Then we describe the performance statistical analysis module from both performance index and threshold sides, and for the performance data, we give some researches on tread prediction and design a trend prediction model based on time series. Finally we provide the design and implementation of the performance statistical analysis function from realtime performance management, historical performance data comparison and performance trend prediction sides, and make it possible to manage the performance of communication network from historical, current and future sides.