摘要
提出一种提升小波尺度自适应非线性预测算子的构造方法。通过相空间重构将剖分信号转换成训练样本,采用基于高斯核函数的支持向量回归机算法进行回归训练,给出所构造预测算子的结构,并说明基于高斯核函数实现最小均方误差原则的机理。通过仿真实验验证用所构建预测算子在故障诊断时具有较好的识别能力和较强的抗噪能力,在信号降噪时信噪比较高、效果良好。
A new method of the construction of scale adaptive nonlinear prediction operator was proposed. The split signal was translated into training sample by the reconstruction method of phase space. Support vector regression （SVR） with the gauss kernel was used to construct prediction operator. The construction of scale adaptive nonlinear prediction operator was given. And the reason for achieving minimize mean squared error （MMSE） was derived. The simulation result shows that this method has good recognition ability and noise immunity at fault diagnosis, and has better SNR at denoising.
出处
《中南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2012年第3期992-996,共5页
Journal of Central South University:Science and Technology
基金
国家高技术研究发展计划("863"计划)项目(2010AA8090514-C)
关键词
提升小波变换
尺度自适应
非线性预测算子
支持向量回归机
最小均方误差原则
lifting scheme
scale adaptive
nonlinear prediction operator
support vector regression （SVR）
minimizemean squared error （MMSE）
