Application of the blind source separation method to feature extraction of machine sound signals

As the result of vibration emission in air, a machine sound signal carries important information about the working condition of machinery. But in practice, the sound signal is typically received with a very low signal-to-noise ratio. To obtain features of the original sound signal, uncorrelated sound signals must be removed and the wavelet coefficients related to fault condition must be retrieved. In this paper, the blind source separation technique is used to recover the wavelet coefficients of a monitored source from complex observed signals. Since in the proposed blind source separation (BSS) algorithms it is generally assumed that the number of sources is known, the Gerschgorin disk estimator method is introduced to determine the number of sound sources before applying the BSS method. This method can estimate the number of sound sources under non-Gaussian and non-white noise conditions. Then, the partial singular value analysis method is used to select these significant observations for BSS analysis. This method ensures that signals are separated with the smallest distortion. Afterwards, the time-frequency separation algorithm, converted to a suitable BSS algorithm for the separation of a non-stationary signal, is introduced. The transfer channel between observations and sources and the wavelet coefficients of the source signals can be blindly identified via this algorithm. The reconstructed wavelet coefficients can be used for diagnosis. Finally, the separation results obtained from the observed signals recorded in a semi-anechoic chamber demonstrate the effectiveness of the presented methods .     

Transformations in machining. Part 1. enhancement of wavelet transformation neural network (WT-NN) combination with a preprocessor

Properly selected transformation methods obtain the most significant characteristics of metal cutting data efficiently and simplify the classification. Wavelet Transformation (WT) and Neural Networks (NN) combination was used to classify the experimental cutting force data of milling operations previously. Preprocessing (PreP) of the approximation coefficients of the WT is proposed just before the classification by using the Adaptive Resonance Theory (ART2) type NNs. Genetic Algorithm (GA) was used to estimate the weights of each coefficient of the PreP. The WT-PreP-NN (ART2) combination worked at lower vigilances by creating only a few meaningful categories without any errors. The WT-NN (ART2) combination could obtain the same error rate only if very high vigilances are used and many categories are allowed.     

Nd3+ Doped Silicate Glass Photonic Crystal Fiber with Random Hole Distributions

The fabrication of one kind of large core area Nd3 doped silicate glass photonic crystal fiber, and demonstration of the fiber's waveguidence properties were reported. This fiber owns a random air hole distribution in the cladding. The measured minimum loss of this kind of fiber is 10 dB·m-1 at 660 nm. These fibers can sustain only a single mode at least over wavelength ranging from 660 nm to 980 nm.     

A Shortcut Method for Estimating the HETP of Structured Packed Distillation Columns at Elevated Pressure

A shortcut model is developed for predicting the HETP of a structured packed distillation column operating at elevated pressure. The proposed model incorporates the geometrical parameters of the packing, physical properties of the vapor and liquid phases, and the hydrodynamics of the two‐phase flow. The proposed model is tested and validated by comparing the predicted results with the present experimental data and some published HETP data. The results show that the proposed model can predict the experimental data with a deviation smaller than 20 %.     

Evaluation of accuracy of measuring intraocular pressure by handheld non-contact applanation tonometer

PURPOSE: To evaluate the accuracy of measuring intraocular pressure by handheld non-contact applanation tonometer. METHOD: 58 patients' (113 eyes) intraocular pressure were measured by Keeler, non-contact tonometer and R 900 Goldmann applanation tonometer and the results of measurement of intraocular pressure by the two kinds of tonometers were compared. RESULT: The mean intraocular pressure measured by non-contact is 16.31 +/- 5.59 mmHg and 17.49 +/- 6.13 mmHg (1 mmHg = 0.1333 kPa) by Goldmann applanation tonometer, respectively. There was no statistical significance to be found (P > 0.05) between the two methods. By linear correlation and regression analysis, a positive correlation was found between the two methods (r = 0.8942, b = 0.8154). CONCLUSION: The handheld non-contact tonometer has the same accuracy and reliability of measurement of intraocular pressure comparing with Goldmann applanation tonometer, and it can be used in glaucoma clinic and screening.     

Monoalkyl Esters of Isoalkylphosphonic Acid as Extractants in Rare Earth Separation

Four monoalkyl esters of isoalkylphosphonic acids have been studied for the chromatographic separation ofrare earths (RE).The relationship between distribution ratios and acid concentration of the eluants forsamarium and ytterbium has been explored.The characteristics of these extractants loaded onto macroreticularpoly (methyl methacrylate) resin beads for the separation of lanthanum,cerium,praseodymium and neodymiumhave been compared.Among the extractants studied,mono(1-methylheptyl) ester of isooctylphosphonic acidwas found to be the best for RE separation.The acidity for separation is lower than any known P-containingextractants for the same purpose.For instance,with 1.73 mol·L~(-1) nitric acid as eluent,thulium,ytterbium andlutetium can be separated quantitatively.The extractant-loaded resins possess high column efficiency with goodkinetic characteristics.The extraction mechanism was also explored.     

Study of miscibility enhancement in poly(styrene‐co‐4‐vinylphenol)/poly(ethylene oxide) blends by the spin labelling method

The electron spin resonance (ESR) spectra of end‐group spin labelled poly(ethylene oxide) (SLPEO) using 2,2,6,6‐tetramethyl‐piperdine‐1‐oxyl nitroxide and its blends with poly(styrene‐co‐4‐vinylphenol) (STVPhs) of different hydroxyl contents were recorded over a wide temperature range. For a blend of SLPEO and pure polystyrene (PS), the ESR spectrum was composed of a single motion component, indicating that PS was immiscible with PEO. For blends composed of SLPEO and different‐hydroxyl‐content STVPhs, two spectral components with different motion rates were observed over a certain temperature range. The difference between the motion rates should be attributed to micro‐heterogeneity in the blends, with the faster rate corresponding to a nitroxide radical motion trapped in the PEO‐rich domain and the slower rate corresponding to a nitroxide radical motion trapped in the STVPh‐rich domain. Variations in the values of a number of the ESR parameters (T_a, T_d and T_50G) and the apparent activation energy (E_a) with hydroxyl content in the blends indicated that the miscibility of the blends increased with increasing hydrogen‐bonding density due to specific interactions between the hydroxyl groups in STVPh and the ether oxygens in PEO. Copyright © 2004 Society of Chemical Industry     

Evolving Fault-Tolerant Neural Networks

In this paper, genetic algorithm is used to help improve the tolerance of feedforward neural networks against an open fault. The proposed method does not explicitly add any redundancy to the network, nor does it modify the training algorithm. Experiments show that it may profit the fault tolerance as well as the generalisation ability of neural networks.