Python program based on VMD variational mode decomposition algorithm
Can be used in time series and other fields
Features: 1. Based on Python
2. The data is read from the excel file, easy to replace
All the complete code, the code that is guaranteed to run can be seen here.
http://t.csdn.cn/obJlC
http://t.csdn.cn/obJlC
! ! ! If the first link cannot be opened, please click on the personal homepage to view my personal introduction.
(After searching for the product, click on the avatar to see all the codes)
Blog of Black Technology Little Potato_CSDN Blog-Deep Learning, Blogger in 32 MCU Field
1. Introduction to the background of VMD variational mode decomposition algorithm
The full name of VMD is Variational Mode Decomposition, which is a new type of time-frequency analysis algorithm, which can be used to decompose and analyze nonlinear and non-stationary signals, such as audio, video, seismic signals and other fields.
The algorithm works by decomposing the signal into modal components and a high-frequency noise component through mathematical optimization. These modal components contribute different frequency band and amplitude information, which can better reveal the characteristics of the signal.
Compared with other time-frequency analysis algorithms, VMD algorithm has the following advantages:
- Unlike the empirical mode decomposition (EMD) algorithm, VMD does not require sliding window technology and is not affected by the selection of basic functions;
- Unlike the wavelet transform algorithm, VMD is not limited by the analysis of fixed frequency bands, and is more effective in removing noise;
- Different from the FFT transform algorithm, the signal components output by VMD have good time-frequency local properties, which are more suitable for analyzing non-stationary signals.
Therefore, the VMD variational mode decomposition algorithm can improve the data analysis ability in the field of signal processing.
2. Summary of advantages of VMD variational mode decomposition algorithm
The advantages of this algorithm are:
- Compared with other time-frequency analysis algorithms, the VMD algorithm does not need sliding window technology and is not affected by the selection of basic functions; it has wider applicability;
- Unlike the wavelet transform algorithm, VMD can better remove noise;
- Different from the FFT transform algorithm, the signal components output by VMD have good time-frequency local properties, which are more suitable for analyzing non-stationary signals.
# 参数设置
K = 3 # EEMD中该参数设置无效，固定为3，改变该参数可能导致绘图出错
target_value = "预测量" # 要预测的列在excel中的名字
if_plot = True # 是否绘制分解图 True绘制，False不绘制，请注意，该功能在Linux子系统，MacOS等环境中可能不可用
# 数据读取
raw_data = pd.read_excel('./data.xlsx')[target_value].values