Coal Thickness Prediction Method Based on VMD and LSTM

• DOI: 10.3390/electronics11020232
• Publication Date: 2022-01-12T14:10:36Z
• Authors: Yaping Huang; Lei Yan; Yan Cheng; Xuemei Qi; Zhixiong Li
• Keywords: Interpolation; Mode (computer interface)

The change in coal seam thickness has an important influence on mine safety and efficient mining. It is very to predict accurately. However, the accuracy of borehole interpolation BP neural network not high. Variational mode decomposition (VMD) strong denoising ability, long short-term memory (LSTM) especially suitable for prediction complex sequences. This paper presents a method using VMD LSTM. Firstly, empirical (EMD) methods are used denoise simple signals, effect verified. Then, wedge-shaped model constructed, seismic forward modeling analysis carried out. results show that based attributes feasible. On basis original 3D data, VMD-LSTM compared with traditional network. proposed this high basically coincides information exposed by existing boreholes. minimum absolute error predicted only 0.08 m, maximum 0.48 m. indicates predicting thickness. can be as new prediction.