Image sticking in liquid crystal display(LCD)is related to the residual direct current(DC)voltage(RDCV)on the cell and the dynamic response of the liquid crystal materials.According to the capacitance change of the li...Image sticking in liquid crystal display(LCD)is related to the residual direct current(DC)voltage(RDCV)on the cell and the dynamic response of the liquid crystal materials.According to the capacitance change of the liquid crystal cell under the DC bias,the saturated RDCV(SRDCV)can be obtained.The response time can be obtained by testing the optical dynamic response of the liquid crystal cell,thereby evaluating the image sticking problem.Based on this,the image sticking of vertical aligned nematic(VAN)LCD(VAN-LCD)with different cell thicknesses(3.8μm and 11.5μm)and different concentrations ofγ-Fe2O3 nanoparticles(0.017 wt.%,0.034 wt.%,0.051 wt.%,0.068 wt.%,0.136 wt.%,0.204 wt.%,and 0.272 wt.%)was evaluated,and the effect of nano-doping was analyzed.It is found that the SRDCV and response time decrease firstly and then increase with the increase of the doping concentration ofγ-Fe2O3 nanoparticles in the VAN cell.When the doping concentration is 0.034 wt.%,theγ-Fe2O3 nanoparticles can adsorb most of the free impurity ions in liquid crystal materials,resulting in 70%reduction in the SRDCV,8.11%decrease in the decay time,and 15.49%reduction in the rise time.The results show that the doping ofγ-Fe2O3 nanoparticles can effectively improve the image sticking of VAN-LCD and provide useful guidance for improving the display quality.展开更多
针对车联网中拒绝服务(denial of service,DoS)攻击难以防范且现有监督学习方法无法有效检测零日攻击的问题,提出了一种混合DoS攻击入侵检测系统.首先,对数据集进行预处理,提高数据的质量;其次,利用特征选择滤除冗余特征,旨在获得代表...针对车联网中拒绝服务(denial of service,DoS)攻击难以防范且现有监督学习方法无法有效检测零日攻击的问题,提出了一种混合DoS攻击入侵检测系统.首先,对数据集进行预处理,提高数据的质量;其次,利用特征选择滤除冗余特征,旨在获得代表性更强的特征;再次,采用集成学习方法将5种基于树结构的监督分类器堆叠集成用于检测已知DoS攻击;最后,提出了一种无监督异常检测方法,将卷积去噪自动编码器与注意力机制相结合来建立正常行为模型,用于检测堆叠集成模型漏报的未知DoS攻击.实验结果表明,对于已知DoS攻击检测,所提系统在Car-Hacking数据集和CICIDS2017数据集上的检测准确率分别为100%和99.967%;对于未知DoS攻击检测,所提系统在上述两个数据集上的检测准确率分别为100%和83.953%,并且在两个数据集上的平均测试时间分别为0.072 ms和0.157 ms,验证了所提系统的有效性和可行性.展开更多
基金supported by the National Natural Science Foundation of China(NSFC)(Nos.11374087 and11504080)the Natural Science Foundation of Hebei Province of China(Nos.A2019202235 and A2017202004)the Key Subject Construction Project of Hebei Province University。
文摘Image sticking in liquid crystal display(LCD)is related to the residual direct current(DC)voltage(RDCV)on the cell and the dynamic response of the liquid crystal materials.According to the capacitance change of the liquid crystal cell under the DC bias,the saturated RDCV(SRDCV)can be obtained.The response time can be obtained by testing the optical dynamic response of the liquid crystal cell,thereby evaluating the image sticking problem.Based on this,the image sticking of vertical aligned nematic(VAN)LCD(VAN-LCD)with different cell thicknesses(3.8μm and 11.5μm)and different concentrations ofγ-Fe2O3 nanoparticles(0.017 wt.%,0.034 wt.%,0.051 wt.%,0.068 wt.%,0.136 wt.%,0.204 wt.%,and 0.272 wt.%)was evaluated,and the effect of nano-doping was analyzed.It is found that the SRDCV and response time decrease firstly and then increase with the increase of the doping concentration ofγ-Fe2O3 nanoparticles in the VAN cell.When the doping concentration is 0.034 wt.%,theγ-Fe2O3 nanoparticles can adsorb most of the free impurity ions in liquid crystal materials,resulting in 70%reduction in the SRDCV,8.11%decrease in the decay time,and 15.49%reduction in the rise time.The results show that the doping ofγ-Fe2O3 nanoparticles can effectively improve the image sticking of VAN-LCD and provide useful guidance for improving the display quality.
文摘针对车联网中拒绝服务(denial of service,DoS)攻击难以防范且现有监督学习方法无法有效检测零日攻击的问题,提出了一种混合DoS攻击入侵检测系统.首先,对数据集进行预处理,提高数据的质量;其次,利用特征选择滤除冗余特征,旨在获得代表性更强的特征;再次,采用集成学习方法将5种基于树结构的监督分类器堆叠集成用于检测已知DoS攻击;最后,提出了一种无监督异常检测方法,将卷积去噪自动编码器与注意力机制相结合来建立正常行为模型,用于检测堆叠集成模型漏报的未知DoS攻击.实验结果表明,对于已知DoS攻击检测,所提系统在Car-Hacking数据集和CICIDS2017数据集上的检测准确率分别为100%和99.967%;对于未知DoS攻击检测,所提系统在上述两个数据集上的检测准确率分别为100%和83.953%,并且在两个数据集上的平均测试时间分别为0.072 ms和0.157 ms,验证了所提系统的有效性和可行性.
