随着强力、高效切削方式的推进以及难加工材料的加工,由切削力引起的工件加工误差(简称切削力误差,cutting force induced error)不容忽视。本文提出切削力误差直接测量法,实现了工作过程中切削力和误差数值的直接对应。通过对不同测量...随着强力、高效切削方式的推进以及难加工材料的加工,由切削力引起的工件加工误差(简称切削力误差,cutting force induced error)不容忽视。本文提出切削力误差直接测量法,实现了工作过程中切削力和误差数值的直接对应。通过对不同测量方法所反映机床误差特性的甄别,提出建立基于"一维"和"二维"测量技术的混合误差模型的思想,研究成果能够为数控机床动态行为对加工性能的影响机理提供试验数据和理论基础。展开更多
针对图像重建过程中噪声去除问题,提出一种自适应加权编码L1/2正则化重建算法。首先,考虑到许多真实图像中不仅含有高斯噪声,而且含有拉普拉斯噪声,设计一种改进的L1-L2混合误差模型(IHEM)算法,该算法兼顾了L1范数与L2范数的各自优点;其...针对图像重建过程中噪声去除问题,提出一种自适应加权编码L1/2正则化重建算法。首先,考虑到许多真实图像中不仅含有高斯噪声,而且含有拉普拉斯噪声,设计一种改进的L1-L2混合误差模型(IHEM)算法,该算法兼顾了L1范数与L2范数的各自优点;其次,由于迭代过程中噪声分布会发生改变,设计一种自适应隶属度算法,该算法可以减少迭代次数和运算时间;利用一种自适应加权编码方法,该方法可以有效地去除含有重尾分布特性的拉普拉斯噪声;另外,设计一种L1/2正则化算法,该算法可以得到较稀疏的解。实验结果表明,相比IHEM算法,自适应L1/2正则化图像重建算法的峰值信噪比(PSNR)平均提高了3.46 d B,结构相似度(SSIM)平均提高了0.02,对含有多种噪声的图像处理具有比较理想的效果。展开更多
Due to the nature of ultra-short-acting opioid remifentanil of high time-varying,complex compartment model and low-accuracy of plasma concentration prediction,the traditional estimation method of population pharmacoki...Due to the nature of ultra-short-acting opioid remifentanil of high time-varying,complex compartment model and low-accuracy of plasma concentration prediction,the traditional estimation method of population pharmacokinetics parameters,nonlinear mixed effects model(NONMEM),has the abuses of tedious work and plenty of man-made jamming factors.The Elman feedback neural network was built.The relationships between the patients’plasma concentration of remifentanil and time,patient’age,gender,lean body mass,height,body surface area,sampling time,total dose,and injection rate through network training were obtained to predict the plasma concentration of remifentanil,and after that,it was compared with the results of NONMEM algorithm.In conclusion,the average error of Elman network is 6.34%,while that of NONMEM is 18.99%.The absolute average error of Elman network is 27.07%,while that of NONMEM is 38.09%.The experimental results indicate that Elman neural network could predict the plasma concentration of remifentanil rapidly and stably,with high accuracy and low error.For the characteristics of simple principle and fast computing speed,this method is suitable to data analysis of short-acting anesthesia drug population pharmacokinetic and pharmacodynamics.展开更多
Machine learning techniques which are about the construction and study of system that can learn from data are combined with many application fields.A method on ionospheric total electron content(TEC)mapping is propose...Machine learning techniques which are about the construction and study of system that can learn from data are combined with many application fields.A method on ionospheric total electron content(TEC)mapping is proposed based on radical basis function(RBF)neural network improved by Gaussian mixture model(GMM).Due to the complicated ionospheric behavior over China,GMM is used to determine the center of basis function in the unsupervised training process.Gradient descent is performed to update the weights function on a sum of squared output error function in the supervised learning process.The TEC values from the center for orbit determination in Europe(CODE)global ionospheric maps covering the period from 2007to 2010 are used to investigate the performance of the developed network model.For independent validation,the simulated TEC values at different latitudes(20°N,30°N and 40°N)along 120°E longitude are analyzed and evaluated.The results show that the simulated TEC from the RBF network based model has good agreement with the observed CODE TEC with acceptable errors.The theoretical research indicates that RBF can offer a powerful and reliable alternative to the design of ionospheric TEC forecast technologies and thus make a significant contribution to the ionospheric modeling efforts in China.展开更多
文摘随着强力、高效切削方式的推进以及难加工材料的加工,由切削力引起的工件加工误差(简称切削力误差,cutting force induced error)不容忽视。本文提出切削力误差直接测量法,实现了工作过程中切削力和误差数值的直接对应。通过对不同测量方法所反映机床误差特性的甄别,提出建立基于"一维"和"二维"测量技术的混合误差模型的思想,研究成果能够为数控机床动态行为对加工性能的影响机理提供试验数据和理论基础。
文摘针对图像重建过程中噪声去除问题,提出一种自适应加权编码L1/2正则化重建算法。首先,考虑到许多真实图像中不仅含有高斯噪声,而且含有拉普拉斯噪声,设计一种改进的L1-L2混合误差模型(IHEM)算法,该算法兼顾了L1范数与L2范数的各自优点;其次,由于迭代过程中噪声分布会发生改变,设计一种自适应隶属度算法,该算法可以减少迭代次数和运算时间;利用一种自适应加权编码方法,该方法可以有效地去除含有重尾分布特性的拉普拉斯噪声;另外,设计一种L1/2正则化算法,该算法可以得到较稀疏的解。实验结果表明,相比IHEM算法,自适应L1/2正则化图像重建算法的峰值信噪比(PSNR)平均提高了3.46 d B,结构相似度(SSIM)平均提高了0.02,对含有多种噪声的图像处理具有比较理想的效果。
基金Project(31200748)supported by the National Natural Science Foundation of China
文摘Due to the nature of ultra-short-acting opioid remifentanil of high time-varying,complex compartment model and low-accuracy of plasma concentration prediction,the traditional estimation method of population pharmacokinetics parameters,nonlinear mixed effects model(NONMEM),has the abuses of tedious work and plenty of man-made jamming factors.The Elman feedback neural network was built.The relationships between the patients’plasma concentration of remifentanil and time,patient’age,gender,lean body mass,height,body surface area,sampling time,total dose,and injection rate through network training were obtained to predict the plasma concentration of remifentanil,and after that,it was compared with the results of NONMEM algorithm.In conclusion,the average error of Elman network is 6.34%,while that of NONMEM is 18.99%.The absolute average error of Elman network is 27.07%,while that of NONMEM is 38.09%.The experimental results indicate that Elman neural network could predict the plasma concentration of remifentanil rapidly and stably,with high accuracy and low error.For the characteristics of simple principle and fast computing speed,this method is suitable to data analysis of short-acting anesthesia drug population pharmacokinetic and pharmacodynamics.
基金supported by the National Natural Science Foundation of China(Grant No.41104096)
文摘Machine learning techniques which are about the construction and study of system that can learn from data are combined with many application fields.A method on ionospheric total electron content(TEC)mapping is proposed based on radical basis function(RBF)neural network improved by Gaussian mixture model(GMM).Due to the complicated ionospheric behavior over China,GMM is used to determine the center of basis function in the unsupervised training process.Gradient descent is performed to update the weights function on a sum of squared output error function in the supervised learning process.The TEC values from the center for orbit determination in Europe(CODE)global ionospheric maps covering the period from 2007to 2010 are used to investigate the performance of the developed network model.For independent validation,the simulated TEC values at different latitudes(20°N,30°N and 40°N)along 120°E longitude are analyzed and evaluated.The results show that the simulated TEC from the RBF network based model has good agreement with the observed CODE TEC with acceptable errors.The theoretical research indicates that RBF can offer a powerful and reliable alternative to the design of ionospheric TEC forecast technologies and thus make a significant contribution to the ionospheric modeling efforts in China.
