继电保护词汇语义识别与应用

A kind of method to recognise vocabulary in grid's protection and control system

  • 摘要: 本文提出了一种电力系统调度、远动、继电保护测试使用的点表自动匹配的算法,该算法有卷积和池化两个步骤,可以对相近词条做一定识别,从而选出最为相近的那一个。在进行语义拆分和加入相似性因子之后,算法可以比其它文献中用到的最小编辑距离法具有更快的速度以及更高的识别率,以此为基础可以实现电力系统通信组态配置和继电保护测试方向上更多智能化的应用。

     

    Abstract: This paper proposes an algorithm for automatic matching of point tables used in power system dispatching, telecontrol, and relay protection testing. This algorithm has two steps of convolution and pooling, which can identify similar entries and select the closest one. After performing semantic splitting and adding similarity factors, the algorithm can have faster speed and higher recognition rate than the minimum edit distance method used in other papers. Based on this, we can achieve more intelligent applications in the direction ofhe power system communication configuration, relay protection testing, etc.

     

/

返回文章
返回