Diagnosis of multiple faults using a double parallel two-hidden-layer extreme learning machine

Multiple faults are easily confused with single faults.In order to identify multiple faults more accurately,a highly efficient learning method is proposed based on a double parallel two-hidden-layer extreme learning machine,called DPTELM.The DPT-ELM method is a variant of an extreme learning machine(ELM).There are some issues with ELM.First,achieving a high accuracy requires too many hidden nodes;second,the direct connection between the input layer and the output layer is ignored.Accordingly,to deal with the above-mentioned problems,DPT-ELM extends the single-hidden-layer ELM to a two-hidden-layer ELM,which can achieve a desired performance with fewer hidden nodes.In addition,a direct connection is built between the input layer and the output layer.Since the input layer weights and the thresholds of the two hidden layers are determined randomly,this simplifies the improved model and shortens the calculation time.Additionally,to improve the signal to noise ratio(SNR),an adaptive waveform decomposition(AWD)algorithm is used to denoise the vibration signal.Then,the denoised signal is used to extract the eigenvalues by the time-domain and frequency-domain methods.Finally,the eigenvalues are input to the DPT-ELM classifier.In this paper,two groups of rolling bearing data at different speeds,which were collected from a real experimental platform,are used to test the method.Each set of data includes three single fault states,two complex fault states and a healthy state.The experimental results demonstrate that the DPT-ELM method achieves fast learning speed and a high accuracy.Moreover,based on 10-fold cross-validation,it proves to be an effective method to improve the accuracy with fewer hidden nodes.

Forest canopy height mapping over China using GLAS and MODIS data

The Geoscience Laser Altimeter System(GLAS)accurately detects the vertical structural information of a target within its laser spot and is a promising system for the inversion of structural features and other biophysical parameters of forest ecosystems.Since the GLAS footprints are discontinuously distributed with a relativity low density,continuous vegetation height distributions cannot be mapped with a high accuracy using GLAS data alone.The MODIS BRDF product provides more forest structural information than other optical remote sensing data.This study aimed to map forest canopy heights over China from the GLAS and MODIS BRDF data.Firstly,the waveform characteristic parameters were extracted from the GLAS data by the method of wavelet analysis,and the terrain index was calculated using the ASTER GDEM data.Secondly,the model reducing the topographic influence was constructed from the waveform characteristic parameters and terrain index.Thirdly,the final canopy height estimation model was constructed from the neural network combining the canopy height estimated with the GLAS point and the MODIS BRDF data,and applied to get the continuous canopy height map over China.Finally,the map was validated by the measured data and the airborne Li DAR data,and the validation results indicated that forest canopy heights can be estimated with high accuracy from combined GLAS and MODIS data.