基于阶段注意力机制的电力线提取算法

电力系统维护是电力系统稳定运行的重要保障，应用智能算法的无人机电力巡检则为电力系统维护提供便捷。电力线提取是自主电力巡检以及保障飞行器低空飞行安全的关键技术，结合深度学习理论进行电力线提取是电力巡检的重要突破点。本文将深度学习方法用于电力线提取任务，结合电力线图像特点嵌入改进的图像输入策略和注意力模块，提出一种基于阶段注意力机制的电力线提取模型（SA-Unet）。本文提出的SA-Unet模型编码阶段采用阶段输入融合策略（Stage input fusion strategy， SIFS），充分利用图像的多尺度信息减少空间位置信息丢失。解码阶段通过嵌入阶段注意力模块（Stage attention module，SAM）聚焦电力线特征，从大量信息中快速筛选出高价值信息。实验结果表明，该方法在复杂背景的多场景中具有良好的性能。     

变压器内部故障在线检测方法仿真及误差分析

为研究变压器内部故障在线检测方法的有效性及抗干扰性,搭建实际变压器的非线性三维有限元模型.对变压器常见几种类型的内部故障进行详细的结构仿真,通过"场-路耦合法"构建各类型故障的ΔV-I1轨迹.同时,通过对正常绕组的测量信号设置一定的误差,构建了考虑测量误差时的误差轨迹.在此基础上,将正常时的标准轨迹、误差轨迹和各种类型的故障轨迹进行对比分析,通过对ΔV-I1轨迹特征差异的定性分析与定量计算,验证该方法在实际应用中的有效性,总结判断各类型故障的规则,同时针对测量误差所造成的干扰提出了建议.     

Preparation and flocculation properties of biodegradable konjac glucomannan-grafted poly(trimethyl allyl ammonium chloride)

Polymer Bulletin - Preparation of konjac glucomannan-grafted poly(trimethyl allyl ammonium chloride) (KGM-g-PTMAAC) was carried out using KGM as polysaccharide matrix and TMAAC as cationic...     

碳脱氧在钢包顶渣改质中的应用

 为了进一步降低夹杂物缺陷并提高产品质量,基于碳脱氧进行了钢包顶渣改质的研究。冷轧产品的生产工艺为铁水预处理→转炉→RH精炼→连铸,为减少钢中夹杂物质量分数,需要进行钢包顶渣改质,同时降低钢包顶渣TFe质量分数。采用粒碳部分替代铝渣球的方法进行基于碳脱氧工艺的钢包顶渣改质,试验结果表明,顶渣改质效果良好,在顶渣TFe质量分数、中间包钢水游离氧明显降低的同时铸坯中Al₂O₃夹杂物得到优化;“30 kg粒渣+铝渣球”工艺降低生产成本5.16元/t(钢)。     

500 kV 开关弹簧机构支架螺栓断裂的原因分析及处理

HPL500B2型弹簧机构支架(BA)螺栓断裂可能导致开关分合闸失败,对其原因进行分析,发现BA螺栓由于预紧固力矩过大,接近其扭断力矩,使得作用在BA螺栓上的力偏高,BA螺栓安全系数仅为12%,安全系数较低。列出BA螺栓更换步骤及注意事项,并详细介绍分闸、合闸弹簧调整步骤,为类似事件的处理提供参考。     

离相式封闭母线工频交流耐压试验补偿装置 在大渡河流域梯级电站中的应用

以某电站封闭母线工频交流耐压为例,根据带防冲击保护电容器的离相式封闭母线(IPB)工频交流耐压试验的方法,通过分析该工频交流耐压补偿试验装置的相关参数计算、选型等重点环节,有效解决了试验过程中出现的问题。     

氟离子选择电极法测定玻璃陶瓷配合料中氟化物的含量

配合料样品溶于水,其中的氟硅酸钠与纯碱反应,形成氟离子,吸取澄清溶液,使氟离子与沉淀物分离,溶液用酸中和过量的碱,并调节pH值到5.5,加入缓冲溶液,定容体积,通过离子选择电极法测定溶液中氟离子的含量。由工作曲线得到每个配合料样品的氟化物含量,计算各个配合料样品氟化物含量数据的相对标准偏差,判断批次配合料的均匀度是否合格。     

Household endotoxin reduction in the Louisa Environmental Intervention Project for rural childhood asthma

Endotoxin exacerbates asthma. We designed the Louisa Environmental Intervention Project (LEIP) and assessed its effectiveness in reducing household endotoxin and improving asthma symptoms in rural Iowa children. Asthmatic school children (N = 104 from 89 homes) of Louisa and Keokuk counties in Iowa (aged 5-14 years) were recruited and block-randomized to receive extensive (education + professional cleaning) or educational interventions. Environmental sampling collection and respiratory survey administration were done at baseline and during three follow-up visits. Mixed-model analyses were used to assess the effect of the intervention on endotoxin levels and asthma symptoms in the main analysis and of endotoxin reduction on asthma symptoms in exploratory analysis. In the extensive intervention group, dust endotoxin load was significantly reduced in post-intervention visits. The extensive compared with the educational intervention was associated with significantly decreased dust endotoxin load in farm homes and less frequent nighttime asthma symptoms. In exploratory analysis, dust endotoxin load reduction from baseline was associated with lower total asthma symptoms score (Odds ratio: 0.52, 95% confidence interval: 0.29-0.92). In conclusion, the LEIP intervention reduced household dust endotoxin and improved asthma symptoms. However, endotoxin reductions were not sustained post-intervention by residents.