In this paper, we propose to generalize the coding schemes first proposed by Kozic &al to high spectral efficient modulation schemes. We study at first Chaos Coded Modulation based on the use of small ...In this paper, we propose to generalize the coding schemes first proposed by Kozic &al to high spectral efficient modulation schemes. We study at first Chaos Coded Modulation based on the use of small dimensional modulo-MAP encoding process and we give a solution to study the distance spectrum of such coding schemes to accurately predict their performances. However, the obtained performances are quite poor. To improve them, we use then a high dimensional modulo-MAP mapping process similar to the low-density generator-matrix codes (LDGM) introduced by Kozic &al. The main difference with their work is that we use an encoding and decoding process on GF (2m) which enables to obtain better performances while preserving a quite simple decoding algorithm when we use the Extended Min-Sum (EMS) algorithm of Declercq &Fossorier.展开更多
函数型聚类分析在统计学领域被广泛关注,其分析过程通常在降维目标实现后进行。为了有效解决函数型主成分聚类问题,文章结合局部线性嵌入算法(Locally Linear Embedding,LLE)在非线性空间下的适用性,提出了一种局部线性下的函数型主成...函数型聚类分析在统计学领域被广泛关注,其分析过程通常在降维目标实现后进行。为了有效解决函数型主成分聚类问题,文章结合局部线性嵌入算法(Locally Linear Embedding,LLE)在非线性空间下的适用性,提出了一种局部线性下的函数型主成分分析模型(LLE Function Principle Component Analysis,LFPCA)。首先,采用函数型主成分分析法作为降维目标方法,改进了FPCA的算法模型,通过将LLE算法的权重系数矩阵与函数型主成分定义相结合,构建出一个适用于非线性空间下的聚类算法;其次,在求解算法的过程中定义了函数型主成分得分,并结合EM算法构建出GMM模型来近似函数型算法的概率密度函数,使模型更高效且适用性更强;最后,通过随机模拟实验及应用分析验证了LFPCA算法模型在真实数据集上具有良好的聚类效能。展开更多
为探究机车车轮退化过程中呈现的两阶段特征问题,提出一种基于两阶段维纳过程的车轮剩余寿命预测方法。利用两阶段维纳过程模型建立车轮轮缘退化模型,通过随机化漂移系数表征车轮退化过程中存在的个体差异;利用期望最大化(expectation m...为探究机车车轮退化过程中呈现的两阶段特征问题,提出一种基于两阶段维纳过程的车轮剩余寿命预测方法。利用两阶段维纳过程模型建立车轮轮缘退化模型,通过随机化漂移系数表征车轮退化过程中存在的个体差异;利用期望最大化(expectation maximum,EM)算法及贝叶斯理论实现了退化模型参数的离线估计与在线更新;通过Schwarz信息准则(Schwarz information criterion,SIC)判断并找到车轮退化过程中存在的变点;最后通过某机车车轮实测轮缘退化数据进行了实例验证。结果表明:与单阶段退化模型相比,考虑存在变点的两阶段退化模型更符合现场实际且在车轮80%寿命分位点处预测精度提升了9.42%。剩余寿命预测结果可以为车轮镟修周期的优化提供一定的理论依据。展开更多
文摘In this paper, we propose to generalize the coding schemes first proposed by Kozic &al to high spectral efficient modulation schemes. We study at first Chaos Coded Modulation based on the use of small dimensional modulo-MAP encoding process and we give a solution to study the distance spectrum of such coding schemes to accurately predict their performances. However, the obtained performances are quite poor. To improve them, we use then a high dimensional modulo-MAP mapping process similar to the low-density generator-matrix codes (LDGM) introduced by Kozic &al. The main difference with their work is that we use an encoding and decoding process on GF (2m) which enables to obtain better performances while preserving a quite simple decoding algorithm when we use the Extended Min-Sum (EMS) algorithm of Declercq &Fossorier.
文摘函数型聚类分析在统计学领域被广泛关注,其分析过程通常在降维目标实现后进行。为了有效解决函数型主成分聚类问题,文章结合局部线性嵌入算法(Locally Linear Embedding,LLE)在非线性空间下的适用性,提出了一种局部线性下的函数型主成分分析模型(LLE Function Principle Component Analysis,LFPCA)。首先,采用函数型主成分分析法作为降维目标方法,改进了FPCA的算法模型,通过将LLE算法的权重系数矩阵与函数型主成分定义相结合,构建出一个适用于非线性空间下的聚类算法;其次,在求解算法的过程中定义了函数型主成分得分,并结合EM算法构建出GMM模型来近似函数型算法的概率密度函数,使模型更高效且适用性更强;最后,通过随机模拟实验及应用分析验证了LFPCA算法模型在真实数据集上具有良好的聚类效能。
文摘为探究机车车轮退化过程中呈现的两阶段特征问题,提出一种基于两阶段维纳过程的车轮剩余寿命预测方法。利用两阶段维纳过程模型建立车轮轮缘退化模型,通过随机化漂移系数表征车轮退化过程中存在的个体差异;利用期望最大化(expectation maximum,EM)算法及贝叶斯理论实现了退化模型参数的离线估计与在线更新;通过Schwarz信息准则(Schwarz information criterion,SIC)判断并找到车轮退化过程中存在的变点;最后通过某机车车轮实测轮缘退化数据进行了实例验证。结果表明:与单阶段退化模型相比,考虑存在变点的两阶段退化模型更符合现场实际且在车轮80%寿命分位点处预测精度提升了9.42%。剩余寿命预测结果可以为车轮镟修周期的优化提供一定的理论依据。
