摘要
为解决缺失数据等条件下配电网的可靠性评估问题,针对配电网可靠性评估时存在评估效果差、计算量大、执行效率低等情况,基于粒子群优化-深度信念网络(PSO-DBN)对配电网可靠性进行分析。首先,设计了基于生成对抗网络(GAN)的电力数据增强模型,从而改善电力数据缺失和不平衡等问题。其次,建立了结合深度信念网络(DBN)和粒子群优化(PSO)模型的优化学习网络,从而得到更准确的配电网可靠性分析结果。以IEEE39电力节点系统为基础,对所提模型进行仿真与分析。仿真结果表明,所提模型性能最优。该研究能够为配电网可靠性评估、管理及稳定运行提供借鉴。
To solve the problem of reliability assessment of distribution networks under conditions such as missing data,the reliability of distribution network is analyzed based on particle swarm optimization-deep belief network(PSO-DBN)in view of the poor assessment effect,large computation,and low execution efficiency in the reliability assessment of distribution network.Firstly,a power data enhancement model based on generative adversarial network(GAN)is designed to improve the problems such as missing and unbalanced power data.Secondly,an optimization learning network combining deep belief network(DBN)and particle swarm optimization(PSO)model is established to obtain more accurate distribution network reliability analysis results.The proposed model is simulated and analyzed based on the IEEE 39 power node system.The simulation results show that the proposed model performs optimally.The research can provide reference for distribution network reliability assessment,management,and stable operation.
作者
张俊成
崔志威
陶毅刚
黎敏
ZHANG Juncheng;CUI Zhiwei;TAO Yigang;LI Min(Guangxi Electricity Power Co.,Ltd.,Nanning 530023,China)
出处
《自动化仪表》
CAS
2024年第5期112-117,共6页
Process Automation Instrumentation
关键词
电力系统
配电网
可靠性评估
深度学习
深度信念网络
粒子群优化
仿真分析
Power system
Distribution network
Reliability assessment
Deep learning
Deep belief network(DBN)
Particle swarm optimization(PSO)
Simulation analysis