With a complex wavelet function, a new real-time recursive algorithm of wavelet transform (WT) is analyzed in detail. Compared with the existing recursive algorithm in two directions, the computing time is greatly red...With a complex wavelet function, a new real-time recursive algorithm of wavelet transform (WT) is analyzed in detail. Compared with the existing recursive algorithm in two directions, the computing time is greatly reduced in response to faults signals in power systems, and the same recursive algorithm can be generalized to other wavelet functions. With the phases and magnitudes of complex WT coefficients under the fast recursive algorithm, a method to detect faults signals of power systems is presented. Lastly, the analyzing results of some signals show that it is effective and practical for the complex wavelet and its real-time recursive algorithm to detect faults of power systems.展开更多
为提高月径流时间序列预测精度,改进最小二乘孪生支持向量回归机(Least Squares Twin Support Vector Regression,LSTSVR)性能,首次基于高斯核函数、多项式核函数、线性核函数构建混合核函数,提出4种季节趋势分解(Seasonal and Trend de...为提高月径流时间序列预测精度,改进最小二乘孪生支持向量回归机(Least Squares Twin Support Vector Regression,LSTSVR)性能,首次基于高斯核函数、多项式核函数、线性核函数构建混合核函数,提出4种季节趋势分解(Seasonal and Trend decomposition using Loess,STL)-小波包变换(Wavelet Packet Transform,WPT)-裂狐优化(Rüppell's Fox Optimizer,RFO)算法-混合核最小二乘孪生支持向量回归机(Hybrid Kernel Least Squares Twin Support Vector Regression,HLSTSVR)模型,并构建STL-WPT-RFO-LSTSVR、STL-WPT-RFO-混合核最小二乘支持向量回归机(Hybrid Kerllel Least Squares Twin Suppart Vector Regression,HLSSVR)、STL-WPT-RFO-最小二乘支持向量回归机(Least Squares Support Vector Regression,LSSVR)等17种对比分析模型,通过云南省高桥、凤屯水文站月径流时间序列预测实例对21种模型进行验证。首先利用STL-WPT二次分解技术对月径流序列进行分解处理,合理划分训练集和验证集;然后基于高斯核函数、多项式核函数、线性核函数,采用“三三”线性组合和“两两”线性组合的方式构建4种混合核函数对月径流分解分量进行空间映射;最后利用RFO寻优HLSTSVR/LSTSVR/HLSSVR/LSSVR最佳超参数,利用最佳超参数建立21种模型对实例月径流序列各分解分量进行训练、预测和重构。结果表明:①4种STL-WPT-RFO-HLSTSVR模型能适应不同尺度的月径流数据分布,具有较好的模型性能和较小的预测误差,其中STL-WPT-RFO-HLSTSVR(高斯+多项式+线性)模型对高桥、凤屯站月径流预测的平均绝对百分比误差MAPE分别为2.85%、2.19%,决定系数R2均为0.9994,预测精度最高、效果最好;②混合核函数兼顾了不同核函数优势,能在模型复杂度与泛化能力之间取得平衡,显著提升模型性能和预测精度;③STL-WPT二次分解技术能有效解决复杂时间序列的非平稳性、非线性和多尺度特征,较STL更具分解优势;④组合模型融合了STL-WPT、RFO和HLSTSVR优点,具有较好的普适性和参考价值。展开更多
为克服传统正交小波变换盲均衡算法(Wavelet transform constant modulus blind equalization algorithm,WTCMA)收敛速度慢、均方误差大、易于陷入局部极小值的缺点,提出了一种基于DNA遗传优化的正交小波常模盲均衡算法(Wavelet transfo...为克服传统正交小波变换盲均衡算法(Wavelet transform constant modulus blind equalization algorithm,WTCMA)收敛速度慢、均方误差大、易于陷入局部极小值的缺点,提出了一种基于DNA遗传优化的正交小波常模盲均衡算法(Wavelet transform constant modulus blind equalization algorithm based on the optimization of DNA genetic algorithm,DNA-GA-WTCMA)。该算法采用基于DNA核苷酸链的编码方式表示问题的可能解,并且对编码后的DNA链采用新型的交叉操作和变异操作来寻找DNA种群中的最优个体,然后将得到的最优个体进行解码,把解码后得到的权向量作为均衡器的最优权向量,以避免WTCMA出现局部收敛并提高收敛速度。仿真实验表明,与基于遗传优化的正交小波变换常模盲均衡算法(Wavelet transform constant modulus blind equalization algorithm based on the optimization of genetic algorithm,GA-WTCMA)相比,该算法可以获得更快的收敛速度和更低的均方误差。展开更多
基金Project supported by Guangdong Province Natural Science Foundation and Central China Electric Power Group Co.
文摘With a complex wavelet function, a new real-time recursive algorithm of wavelet transform (WT) is analyzed in detail. Compared with the existing recursive algorithm in two directions, the computing time is greatly reduced in response to faults signals in power systems, and the same recursive algorithm can be generalized to other wavelet functions. With the phases and magnitudes of complex WT coefficients under the fast recursive algorithm, a method to detect faults signals of power systems is presented. Lastly, the analyzing results of some signals show that it is effective and practical for the complex wavelet and its real-time recursive algorithm to detect faults of power systems.
文摘为提高月径流时间序列预测精度,改进最小二乘孪生支持向量回归机(Least Squares Twin Support Vector Regression,LSTSVR)性能,首次基于高斯核函数、多项式核函数、线性核函数构建混合核函数,提出4种季节趋势分解(Seasonal and Trend decomposition using Loess,STL)-小波包变换(Wavelet Packet Transform,WPT)-裂狐优化(Rüppell's Fox Optimizer,RFO)算法-混合核最小二乘孪生支持向量回归机(Hybrid Kernel Least Squares Twin Support Vector Regression,HLSTSVR)模型,并构建STL-WPT-RFO-LSTSVR、STL-WPT-RFO-混合核最小二乘支持向量回归机(Hybrid Kerllel Least Squares Twin Suppart Vector Regression,HLSSVR)、STL-WPT-RFO-最小二乘支持向量回归机(Least Squares Support Vector Regression,LSSVR)等17种对比分析模型,通过云南省高桥、凤屯水文站月径流时间序列预测实例对21种模型进行验证。首先利用STL-WPT二次分解技术对月径流序列进行分解处理,合理划分训练集和验证集;然后基于高斯核函数、多项式核函数、线性核函数,采用“三三”线性组合和“两两”线性组合的方式构建4种混合核函数对月径流分解分量进行空间映射;最后利用RFO寻优HLSTSVR/LSTSVR/HLSSVR/LSSVR最佳超参数,利用最佳超参数建立21种模型对实例月径流序列各分解分量进行训练、预测和重构。结果表明:①4种STL-WPT-RFO-HLSTSVR模型能适应不同尺度的月径流数据分布,具有较好的模型性能和较小的预测误差,其中STL-WPT-RFO-HLSTSVR(高斯+多项式+线性)模型对高桥、凤屯站月径流预测的平均绝对百分比误差MAPE分别为2.85%、2.19%,决定系数R2均为0.9994,预测精度最高、效果最好;②混合核函数兼顾了不同核函数优势,能在模型复杂度与泛化能力之间取得平衡,显著提升模型性能和预测精度;③STL-WPT二次分解技术能有效解决复杂时间序列的非平稳性、非线性和多尺度特征,较STL更具分解优势;④组合模型融合了STL-WPT、RFO和HLSTSVR优点,具有较好的普适性和参考价值。
文摘为克服传统正交小波变换盲均衡算法(Wavelet transform constant modulus blind equalization algorithm,WTCMA)收敛速度慢、均方误差大、易于陷入局部极小值的缺点,提出了一种基于DNA遗传优化的正交小波常模盲均衡算法(Wavelet transform constant modulus blind equalization algorithm based on the optimization of DNA genetic algorithm,DNA-GA-WTCMA)。该算法采用基于DNA核苷酸链的编码方式表示问题的可能解,并且对编码后的DNA链采用新型的交叉操作和变异操作来寻找DNA种群中的最优个体,然后将得到的最优个体进行解码,把解码后得到的权向量作为均衡器的最优权向量,以避免WTCMA出现局部收敛并提高收敛速度。仿真实验表明,与基于遗传优化的正交小波变换常模盲均衡算法(Wavelet transform constant modulus blind equalization algorithm based on the optimization of genetic algorithm,GA-WTCMA)相比,该算法可以获得更快的收敛速度和更低的均方误差。
