返回
Distribution network reliability assessment and optimization method based on convolutional neural network
DOI:10.1088/2631-8695/ae0f08.png)
摘要
En 中文
考虑到影响电力配电网运行的各种复杂因素以及传统可靠性评估与优化方法的局限性,本研究提出了一种基于卷积神经网络的新型方法。该模型使用IEEE 123节点配电系统的数据进行了训练和测试,并将其实验性能与多种成熟方法进行了比较,包括故障树分析、支持向量机模型、深度信念网络和随机森林算法。实验结果表明,在系统平均停电持续时间指数方面,所提出的模型实现了每用户平均停电时长3.15小时,显著低于故障树分析模型的每用户5.42小时以及支持向量机模型的每用户4.96小时。在预测性能方面,该模型的准确率达到92.5%,召回率为90.8%,F1得分为91.6%,优于所有对比模型。该方法为评估和优化电力配电网的可靠性提供了一种新颖且有效的方式,从而增强了运行稳定性、成本效益以及电力供应的整体可靠性。
Keyword:
convolutional neural network
reliability assessment
optimization method
power system
期刊
E
IF:
1.6
论文数:
2.1K
被引数:
0
机构
暂无机构信息
引用论文
A novel reliability assessment method for distribution networks based on linear programming considering distribution automation and distributed generation一种基于线性规划,考虑配电自动化和分布式发电的配电网可靠性评估新方法
Reliability assessment of modern distribution networks embedded with renewable and distributed resources嵌入可再生和分布式资源的现代配电网的可靠性评估
Reliability Assessment of Cyber-Physical Distribution System Using Multi-Dimensional Information Network Model基于多维信息网络模型的信息-物流系统可靠性评估
Reliability Assessment of Power Distribution Cyber–Physical Systems Considering the Impact of Wireless Access Networks考虑无线接入网络影响的电力分配网络物理系统可靠性评估
Development and Integration of Momentary Event Models in Active Distribution System Reliability Assessment主动配电系统可靠性评估中瞬时事件模型的开发与集成
Reliability Assessment of Distribution Network Considering Mobile Energy Storage Vehicles and Dynamic Zonal Coupling考虑移动储能车辆和动态分区耦合的配电网可靠性评估
Improved Bootstrap Method Based on RBF Neural Network for Reliability Assessment基于RBF神经网络的改进自助法可靠性评估

