Real-time power system topology calculation method based on graphic calculation

Abstract

The invention relates to a real-time power system topology calculation method based on graphic calculation and belongs to the technical field of power system operation and scheduling. The method comprises the following steps: combining the latest proposed graphic calculation technology in an internet domain, fully considering a calculation characteristic that the node serves as a center, transforming topology analysis into a minimum propagation process, and by virtue of the design of a data structure stored in the nodes and an information transfer manner between the nodes, enabling calculationof each node to only depend on information of the own node, an adjacent node and edge. According to the method, a transformation mechanism of the nodes between an active state and a dormant state isdesigned, and redundant calculation is reduced to the greatest degree. According to the method disclosed by the invention, the node calculation parallelizability and calculation efficiency are improved, and the memory usage problem caused in the original algorithm due to function recursion and retrospect is solved. The method can be used for online analytical application of an energy management system and has significances on real-time optimal scheduling of the power system and online analysis of the safety and reliability of the power system.

Description

technical field [0001] The invention relates to a real-time calculation method of power system topology based on graph calculation, and belongs to the technical field of power system operation scheduling. Background technique [0002] When the power system is in normal operation, many functions of the EMS (Energy Management System) online analysis program are completed based on the node admittance matrix. When a switch or a switch is switched on or off in a power system, the topology of the power grid will change, resulting in a change in the node admittance matrix. If the wiring cannot be modified quickly and accurately according to the change of the switch or knife switch state to form a new node admittance matrix, then the original node admittance matrix cannot reflect the actual system in real time, which will lead to erroneous analysis with judgment. Therefore, according to the real-time status of the switch, quickly and accurately determine the network connection sta...

AI Technical Summary

Problems solved by technology

The traditional distributed computing framework Mapreduce will generate a lot of I / O read and write and node communication overhead when dealing with such complex correlation problems, and the calculation efficiency is not high. For example, some companies use Mapreduce to recommend social users. For 50 million Registered users, 5 billion relationship pairs, using a cluster of 10 machines, requires more than 10 hours of calculation