基于关联向量机的电网有功控制指令执行效果辨识研究  

Research on Identification of Active Power Control Instruction Execution Effect in Power Grid Based on Correlation Vector Machine

作　　者：陈光宇 南钰 赵文贺 袁野[3] CHEN Guangyu;NAN Yu;ZHAO Wenhe;YUAN Ye(Kaifeng Power Supply Company State Grid Henan Electric Power Company,Kaifeng 475000,China;School of Electric Power Engineering Nanjing Institute of Technology,Nanjing 211167,China;School of Electrical Information Engineering,Jiangsu University,Zhenjiang 212013,China)

机构地区：[1]国网河南省电力公司开封供电公司,河南开封475000 [2]南京工程学院电力工程学院,南京211167 [3]江苏大学电气信息工程学院,江苏镇江212013

出　　处：《电工材料》2024年第6期87-90,共4页Electrical Engineering Materials

摘　　要：随着“源-荷”不确定性的增强,电网的控制难度逐渐增大,自动发电控制(AGC)指令执行效果的重要性日益突出,为有效评估指令执行结果的有效性,提出一种基于关联向量机的电网有功控制指令执行效果辨识方法。该方法采用变分自编码器和全连接网络构建有功指令执行效果辨识模型,并在所提模型的基础上,将关联向量机作为输出层进行融合建模,实现了对预测结果在给定偏差范围内的可信度分析,从概率角度增强了辨识结果的使用价值。With the increasing uncertainty of source and load,the control difficulty of the power grid gradually increases,and the importance of the execution effect of active power control(AGC)instructions in the power grid is becoming increasingly prominent.In order to effectively evaluate the effectiveness of instruction execution results,a method for identifying the execution effect of active power control instructions in the power grid based on correlation vector machines was proposed.A variational autoencoder and fully connected network was used to construct an identification model for the execution effect of active power instructions.Based on the proposed model,the correlation vector machine was used as the output layer for fusion modeling,achieving reliability analysis of the predicted results within the given deviation range,and enhancing the value of the identification results from a probabilistic perspective.

关 键 词：关联向量机 有功控制 效果辨识 

分 类 号：TM712[电气工程—电力系统及自动化]