Data sharing mode of dispatching automation system based on distributed machine learning

Abstract

AbstractThe difficulties of sending massive amounts of data between several data centres are examined in this work, with particular attention paid to how poorly current scheduling algorithms handle point‐to‐multipoint transfers and transmission time limits. In this research, a new method called multicast source‐based tree (MSBT) is proposed for effectively handling point‐to‐multipoint transmissions in a certain amount of time. By allowing receivers to simultaneously receive data from several source points, MSBT introduces the idea of ‘source selection’ for the creation of multicast tree structure‐based algorithms. Large data blocks are distributed as efficiently as possible using this method, which also guarantees effective transmission from a single‐source point to several recipient locations. Furthermore covered in the article is how PV producers and sellers' capacity allocation is affected by the discount rate. These results offer insightful information on how decisions are made in related sectors. The development of new energy big data platforms underscores their significance; leaders in the industry, like United Power, Vision Energy and Goldwind, serve as prime examples.