针对特高压接入的输电网网架错综复杂、远景适应性差,甚至出现网架规划与开环分区决策相互独立导致资源浪费等问题,该文引入奇异值分解(singular value decomposition,SVD)和特征索引技术,提出了一种考虑电磁环网开环分区和500/220kV典...
详细信息
针对特高压接入的输电网网架错综复杂、远景适应性差,甚至出现网架规划与开环分区决策相互独立导致资源浪费等问题,该文引入奇异值分解(singular value decomposition,SVD)和特征索引技术,提出了一种考虑电磁环网开环分区和500/220kV典型供电结构的输电网多阶段规划方法。首先,提取典型供电结构的路径特征,制定了一种基于标签路径特征的典型供电结构索引策略;然后,构建了一种基于SVD的典型供电结构相似度指标,基于此,以全局典型供电结构相似程度最大为目标,建立了相似度识别规划模型;最后,提出了一种反馈调节机制及求解方法,统筹协调各阶段网架规划和开环分区之间的决策信息。与传统规划方法相比,该文方法不仅能够避免不同阶段开环分区决策导致的资源浪费,还可以减少复杂的规划评估指标计算,构建供电可靠性高、运行方式灵活、远景适应性强的网架结构。基于湖南某地区实际电网算例验证了该文方法的准确性和有效性。
针对当前有载变压器直流电阻测试中需要专人手动调节档位的问题,本文设计并实现了一种自动化有载变压器直流电阻测试系统,能够自动识别和调节分接开关的档位。系统引入了图像采集与处理模块,结合先进的YOLOv8算法,实现了对分接开关档位的精确捕捉和自动识别,采用香橙派作为核心控制器,负责整体调度与数据交换,确保了测试流程的自动化和智能化。实际应用表明,该系统能够有效提高测试的自动化水平,减少人为操作失误,显著提高了测试效率和准确性。In response to the issue of requiring a dedicated person to manually adjust the gear in the current on load transformer DC resistance testing, this paper designs and implements an automated on load transformer DC resistance testing system that can automatically identify and adjust the gear of the tap changer. The system introduces an image acquisition and processing module, combined with advanced YOLOv8 algorithm, to achieve precise capture and automatic recognition of tap changer gears. Orange Pi is used as the core controller, responsible for overall scheduling and data exchange, ensuring the automation and intelligence of the testing process. Practical application has shown that the system can effectively improve the automation level of testing, reduce human operational errors, and significantly improve testing efficiency and accuracy.
暂无评论