Aiming at the problem that traditional sampling was prone to data redundancy and limited to the processing of static data,this paper proposed a dynamic over sampling method based on manifold ranking.The method used the manifold structure to described data according to the data within the global manifold structure sorting the minority class data.It selected higher sort value to perform a resampling strategy in order to improve the data's equilibrium.The experimental results show that the proposed method can obtain better classification performance than the current similar method,and can effectively improve the recognition performance of the classifier to a small number of classes.

Aiming at the problem that the transformer condition evaluation factors have fuzzy and grey characteristics and fixed weight (FW) can not accurately evaluate the condition,we proposed a condition evaluation method for transformer based on grey fixed weight clustering and variable weight model.The method is applied to evaluate transformer condition layer by layer.Firstly,the confidence degree of the association rules is introduced to determine the FW of key indicators.According to the classification of transformer state grade,the triangle whiten function (TWF) is established.The grey clustering coefficient (GCC) matrix and fault layer evaluation results are obtained through grey clustering (GC) method.Then the variable weight (VW) is obtained by combining variable weight synthesis mode,and then the transformer condition is evaluated by fuzzy evaluation method.The transformer test report data is carried out as the case analysis and results show that the method we proposed can assess the condition of the transformer objectively and accurately.

针对电力通信系统光纤线路未来状态趋势预测问题,提出一种基于自回归移动平均模型(Autoregressive Integrated Moving Average Model,ARIMA)-支持向量机(Support Vector Machine,SVM)组合模型的光功率趋势预测法.根据光功率数据的非线性和时变性特点,首先利用小波变换将其分解并重构,然后设计一种基于混合核函数的SVM预测模型,对重构后的数据分别构建ARIMA模型与基于混合核函数的SVM模型并预测,最后组合2种模型的预测结果实现光功率趋势预测.分别采用ARIMA模型、径向基函数(Radical Basis Function,RBF)模型、SVM模型、ARIMA-RBF模型与该模型进行实验并对比,结果表明,基于混合核函数的ARIMA-SVM组合模型对光功率趋势预测最为准确,其预测精度及计算速度都有了显著提高,为未来光功率趋势发展提供了一种有效可行的预测新方法.

Objective To investigate the protective effect of hydroxytyrosol(HT) on the oxidative damage of HEK-293 cell induced by H2O2.Method As target cells,HEK-293 cells were divided into 5 groups including control group,H2O2-injured group(20 μM),low-dosage of HT group(12.5 μM),mid-dosage of HT group(25 μM) and high-dosage of HT group(50 μM).8-hydroxyderoxyguanosine(8-OHdG) was measured by immunocytochemistry staining analysis,which was a reliable marker for oxidative DNA damage.Reactive oxygen species(ROS) was detected by 2,7-dichlorofluorescein diacetate(DCFH-DA) as a fluorescent probe.Fluorometric methods was used to detect the level of glutathione(GSH).Result Compared with the control group,H2O2 caused significant oxidative damage through 8OHdG formation in HEK-293 cells.The formation of intracellular ROS was significantly increased and GSH was decreased(P < 0.05,P < 0.01).HT treatments significantly decreased 8-OHdG activation and ROS,and increased the level of GSH.Conclusion The results showed that antioxidant HT could protect HEK-293 cells from H2O2-induced oxidative damage.

In order to reduce the cost and increase the recovery rate of residual molten steel,The Second Steelmaking Plant of QingdaoSteel improved turning slag process for producing 82 level cord steel by analyzing the slag system,calculating the sulfur capacity and reducing slag,and pouring out 0.775 t slag from slag pot and adding 0.375 t lime.After improvement annual recoverable quantity of residual molten steel is about 0.018 million t,reduced discharge quantity of slag is 0.015 million t and electric energy saving is 6.2 million kW · h.

应用CCGI对安徽省2010年棉花生产进行评述,与2009年相比,棉花种植面积增3.6%,单产减10.4%,总产减7.1%。棉价一路飙升,且创历史新高,出现减产增收的局面。面积增加主要受2009年棉价上涨的影响。单产下降主要原因是气候异常、灾害天气频发,不利于棉花生长。