基于BiLSTM-CRF 与BT 防误智能调控一体化平台设计

打开文本图片集

中图分类号：TP393；TM734 文献标志码：A 文章编号：1001-5922（2025）06-0184-05

Abstract：In view of the various drawbacks of the traditional power generation dispatching mode，which is difficult to meet the current dispatching requirements，the manual audit of work tickets in the dispatching process has the problems of low efficiency and high risk. Based on the research of Bi-directional Long Short-Term Memory（BiL⁃ STM）-Conditional Random Fields（CRF）entity recognition and Behavior Tree（BT）anti-error intelligent anti-error algorithm，this paper proposes an intelligent anti-error control integrated platform design scheme，gives the platform function framework，and realizes the successful application in Wujiang. Through the test of entity recognition mod⁃ els based on various different algorithms，the verification results show that the entity recognition model based on BiLSTM-CRF algorithm is superior；based on the research of the integrated platform of intelligent anti-misoperation control，the integrated platform of Wujiang regulation and control was developed in Wujiang hydropower station and passed the test，which verified the feasibility and effectiveness of the design scheme and the intelligent anti-misoper⁃ ation function.

Key words：behavior tree；bidirectional long short term memory network；conditional random field；intelligent error prevention；Integration of regulation and control；J2EE

随着电力系统的扩建与复杂化，国网公司提出和实施了大运行、大检修体系[1]，传统调度管理模式的诸多弊病便逐渐暴露[2]。（剩余6537字）