现有基于神经网络的电池荷电状态(state of charge,SOC)预测研究大多把重点放在模型结构和相关参数的优化上,却忽略了训练数据的重要作用.针对该问题,文中提出了一种基于特征选择和数据增强的电池SOC预测方法.首先,方法根据原始电池充...现有基于神经网络的电池荷电状态(state of charge,SOC)预测研究大多把重点放在模型结构和相关参数的优化上,却忽略了训练数据的重要作用.针对该问题,文中提出了一种基于特征选择和数据增强的电池SOC预测方法.首先,方法根据原始电池充放电数据进行特征工程,并使用排列重要性(permutation importance,PI)方法选出对模型预测最有帮助的7个特征;其次,通过加入高斯噪声来扩大训练数据样本总量,达到数据增强的目的.实验使用双向长短时记忆网络(bidirectional long short-term memory,Bi-LSTM)作为预测模型,使用Panasonic 18650PF数据集作为训练数据.使用标准Bi-LSTM进行预测时,平均绝对误差(mean absolute error,MAE)和最大误差(max error,MaxE)分别为0.65%和3.92%,而在进行特征选择和数据增强后,模型预测的MAE和MaxE分别为0.47%和2.62%,表明PI特征工程与高斯数据增强方法可以进一步提升电池荷电状态预测模型的精度.展开更多
目的:探讨CD44与CD33在口腔黏膜良性淋巴组织增生病(benign lymphoadenosis of oral mucosa,BLOM)中的表达及临床意义。方法:选择2017年1月—2020年3月青岛市中医医院病理科77例BLOM蜡块作为实验组,另取同时间段63例正常口腔黏膜组织蜡...目的:探讨CD44与CD33在口腔黏膜良性淋巴组织增生病(benign lymphoadenosis of oral mucosa,BLOM)中的表达及临床意义。方法:选择2017年1月—2020年3月青岛市中医医院病理科77例BLOM蜡块作为实验组,另取同时间段63例正常口腔黏膜组织蜡块作为对照组。采用免疫组织化学法检测2组CD44、CD33阳性表达情况,采用Spearman分析BLOM患者病变组织中CD33与CD44阳性表达的相关性。收集患者一般资料,分析BLOM患者病变组织中CD33、CD44表达与临床病理特征的关系。采用SPSS 21.0软件包对数据进行统计学分析。结果:对照组、实验组CD33阳性表达率分别为95.24%、63.64%,差异有统计学意义(P<0.05);CD44阳性表达率分别为93.65%、67.53%,差异有统计学意义(P<0.05)。Spearman分析结果显示,BLOM患者病变组织中CD33与CD44阳性表达呈正相关(r=0.834,P=0.002);CD33、CD44表达与临床分型、炎症程度、有无淋巴滤泡、淋巴细胞浸润有关(P<0.05),而与年龄、性别、病程、病变部位、上皮表面角化无关(P>0.05)。结论:BLOM患者病变组织中CD33、CD44阳性表达率下降,与临床分型、炎症程度、有无淋巴滤泡、淋巴细胞浸润密切相关。展开更多
When complex networks describe a wide range of systems in nature and society,it is increasingly recognized that the topology of real networks are governed by robust organizing principles.Here we discuss the structural...When complex networks describe a wide range of systems in nature and society,it is increasingly recognized that the topology of real networks are governed by robust organizing principles.Here we discuss the structural metrics such as average path length,clustering coefficient and degree distribution,the main models covering random graphs,small-world and scale-free networks,the interplay between structural properties and the synchronization of complex networks.展开更多
基金supported in part by the Program for New Century Excellent Talents in University of China(No.NCET-06-0510)National Natural Science Foundation of China (No. 60874091)the Natural Science Basic Research Project for Universities of Jiangsu Province(No. 08KJD510022)
文摘When complex networks describe a wide range of systems in nature and society,it is increasingly recognized that the topology of real networks are governed by robust organizing principles.Here we discuss the structural metrics such as average path length,clustering coefficient and degree distribution,the main models covering random graphs,small-world and scale-free networks,the interplay between structural properties and the synchronization of complex networks.
