论文部分内容阅读
经验模式分解(EMD)能够有效获得非平稳非线性信号的时频特征,但传统的EMD分解算法存在严重的端点效应.在深入研究和分析EMD算法的基础上,提出了一种基于波形匹配的端点效应处理方案,通过计算波形匹配度,在平均包络线内部寻找与其端部变化趋势最为接近的子波,并用这段子波代替平均包络线的边缘部分,使处理后的平均包络线极大地接近真实包络线,并把这种端点效应处理方案的EMD分解算法应用到实际的股票市场价格趋势分解中.实验结果表明,与经典的EMD边界延拓算法相比,本文提出的算法能更有效地抑制EMD分解时的边界效应,分解得到的固有模式函数更能体现模拟信号真实的频率、幅值信息.应用实验表明:与现有方法相比,该方法更能提高预测精度.
Empirical Mode Decomposition (EMD) can effectively obtain the time-frequency characteristics of non-stationary nonlinear signals, but the traditional EMD decomposition algorithm has serious endpoint effects.Based on the further study and analysis of the EMD algorithm, a new algorithm based on waveform matching Endpoint effect processing scheme, by calculating the waveform matching degree, find the end of the average envelope within the closest trend of the wavelet, and use this wavelet instead of the average envelope of the edge part of the average envelope after processing Greatly approaching the real envelope, and applies the EMD decomposition algorithm of this endpoint effect solution to the actual price trend decomposition of the stock market.The experimental results show that, compared with the classical EMD boundary extension algorithm, the proposed algorithm Can effectively restrain the boundary effect when EMD is decomposed, and the intrinsic mode function obtained by decomposition can better reflect the real frequency and amplitude information of the analog signal.Experiments show that this method can improve the prediction accuracy better than the existing methods.