为了解决冲击噪声下长短时记忆(long short term memory,LSTM)神经网络调制信号识别方法抗冲击噪声能力弱和超参数难以确定的问题,本文提出了一种演化长短时记忆神经网络的调制识别方法。利用基于短时傅里叶变换的卷积神经网络(convolut...为了解决冲击噪声下长短时记忆(long short term memory,LSTM)神经网络调制信号识别方法抗冲击噪声能力弱和超参数难以确定的问题,本文提出了一种演化长短时记忆神经网络的调制识别方法。利用基于短时傅里叶变换的卷积神经网络(convolution neural network,CNN)去噪模型对数据集去噪;结合量子计算机制和旗鱼优化器(sailfish optimizer,SFO)设计了量子旗鱼算法(quantum sailfish algorithm,QSFA)去演化LSTM神经网络以获得最优的超参数;使用演化长短时记忆神经网络作为分类器进行自动调制信号识别。仿真结果表明,采用所设计的CNN去噪和演化长短时记忆神经网络模型,识别准确率有了大幅度的提高。量子旗鱼算法演化LSTM神经网络模型降低了传统LSTM神经网络容易陷于局部极小值或者过拟合的概率,当混合信噪比为0 dB,所提方法对11种调制信号的平均识别准确率达到90%以上。展开更多
Objective:To investigate the pharmacokinetics of clindamycin phosphate vaginal gel in healthy Chinese female volunteers.Methods:Ten healthy Chinese female volun- teers were intravaginally given with 5.0g of clindamyci...Objective:To investigate the pharmacokinetics of clindamycin phosphate vaginal gel in healthy Chinese female volunteers.Methods:Ten healthy Chinese female volun- teers were intravaginally given with 5.0g of clindamycin phosphate vaginal gel (equivalent to 100mg of clindamycin) once for single dose treatment,and 5.0g,once a day for 3 days,for mul- tiple dose treatment.The serum concentration of clindamycin were determined by HPLC-MS method and its pharmacokinetic parameters of clindamycin were calculated by DAS 1.0 soft- ware.Results:The main pharmacokinetic parameters of clindamycin for single dose and multiple doses were as follows:t_(1/2) were (15.30±2.62) hours and (14.78±2.49) hours,Tmax were (4.88±0.94) hours and (4.70±0.59) hours,Cmax were (38.30±22.77) ng/ml and (44.87±26.71) ng/ml,AUC0_(-∞) were (783.45±351.19) ng·ml^(-1)·h^(-1) and (1015.68±456.95) ng·ml^(-1)·h^(-1),respectively.Conclusion:The Cmax of clindamycin phosphate vaginal gel after a single dose and multiple doses are obviously lower and t_(1,2) are longer than that of clindamycin phosphate oral preparations,which suggests that clindamycin phosphate vaginal gel acts locally and can be slowly absorbed to circulation for systemic actions.展开更多
文摘为了解决冲击噪声下长短时记忆(long short term memory,LSTM)神经网络调制信号识别方法抗冲击噪声能力弱和超参数难以确定的问题,本文提出了一种演化长短时记忆神经网络的调制识别方法。利用基于短时傅里叶变换的卷积神经网络(convolution neural network,CNN)去噪模型对数据集去噪;结合量子计算机制和旗鱼优化器(sailfish optimizer,SFO)设计了量子旗鱼算法(quantum sailfish algorithm,QSFA)去演化LSTM神经网络以获得最优的超参数;使用演化长短时记忆神经网络作为分类器进行自动调制信号识别。仿真结果表明,采用所设计的CNN去噪和演化长短时记忆神经网络模型,识别准确率有了大幅度的提高。量子旗鱼算法演化LSTM神经网络模型降低了传统LSTM神经网络容易陷于局部极小值或者过拟合的概率,当混合信噪比为0 dB,所提方法对11种调制信号的平均识别准确率达到90%以上。
文摘Objective:To investigate the pharmacokinetics of clindamycin phosphate vaginal gel in healthy Chinese female volunteers.Methods:Ten healthy Chinese female volun- teers were intravaginally given with 5.0g of clindamycin phosphate vaginal gel (equivalent to 100mg of clindamycin) once for single dose treatment,and 5.0g,once a day for 3 days,for mul- tiple dose treatment.The serum concentration of clindamycin were determined by HPLC-MS method and its pharmacokinetic parameters of clindamycin were calculated by DAS 1.0 soft- ware.Results:The main pharmacokinetic parameters of clindamycin for single dose and multiple doses were as follows:t_(1/2) were (15.30±2.62) hours and (14.78±2.49) hours,Tmax were (4.88±0.94) hours and (4.70±0.59) hours,Cmax were (38.30±22.77) ng/ml and (44.87±26.71) ng/ml,AUC0_(-∞) were (783.45±351.19) ng·ml^(-1)·h^(-1) and (1015.68±456.95) ng·ml^(-1)·h^(-1),respectively.Conclusion:The Cmax of clindamycin phosphate vaginal gel after a single dose and multiple doses are obviously lower and t_(1,2) are longer than that of clindamycin phosphate oral preparations,which suggests that clindamycin phosphate vaginal gel acts locally and can be slowly absorbed to circulation for systemic actions.
