查看更多>>摘要：? 2022 Elsevier LtdThe bearing acoustic signal is interfered by reflected sounds and background noises, resulting in a low signal-to-noise ratio (SNR). To address this problem, this paper proposes a feature enhancement method that combines recursive least squares (RLS) with resonance-based sparse signal decomposition (RSSD) into the RLS-RSSD method. First, the RLS method is used as the inverse filter to remove the reverberation as well as reduce the interference of the late reflected sound on the direct signal, then RSSD and wavelet denoising are used to eliminate aperiodic component in the low and high frequency bands. The signals are synthesized based on the amplitudes of different frequency signals, and finally, the bearing fault is determined by envelope spectrum analysis. The results of the simulation data, experimental data, and field application data analysis indicate that the frequency of bearing defects can be accurately extracted by the proposed method.
查看更多>>摘要：? 2022 Elsevier LtdFirst, this study develops an enhanced slime mould algorithm (MGSMA). The main idea is to combine the new movement strategy and Gaussian kernel probability strategy to improve the optimization performance of the original slime mould algorithm. These two tactics increase MGSMA's capacity to avoid being stuck in a local optimum and reduce the probability of delaying the convergence process. Second, by integrating non-local mean, 2D Kapur's entropy, and other relevant methodologies, a novel multi-level image segmentation (MLIS) model is developed based on the suggested MGSMA. To showcase MGSMA's performance, specific comparative tests based on IEEE CEC2014 are carried out, clearly showing that MGSMA is a swarm intelligence approach capable of jumping out of the local optimum and the convergence process does not willingly interrupt. To demonstrate that the MGSMA-based MLIS approach can provide high-quality segmentation results, it is compared to eight other comparable methods at both high and low thresholding levels, with some relevant experimental findings to back up its claims. As a consequence, there is no question that MGSMA is a high-performance swarm intelligence optimization approach and that the MGSMA-based MLIS method can provide high-quality segmentation results. The source codes of the SMA algorithm and latest updates are publicly available at https://aliasgharheidari.com/SMA.html.
查看更多>>摘要：? 2022Denoising is commonly used to improve the signal-to-noise ratio of ground-penetrating radar (GPR) for target detection and recognition. We propose an adaptive GPR denoising method based on the fast independent component analysis (FastICA) with wavelet transform modulus maxima (WTMM) multifractal spectrum, which can effectively separate the information of the abnormal body in the reservoir that is submerged by the noise signal. The target signal and background noise signal are extracted with FastICA and identified with multifractal spectrum. The numerical example show that FastICA can effectively separate the components of the target signal and noise signal when the useless signals submerge the target signal. The WTMM multifractal spectrum can solve the disorder problem in the results of FastICA. The results based on the field measured data show that the proposed denoising method has higher stability and convenience than the traditional denoising method, which validates the effectiveness of the proposed method.
查看更多>>摘要：? 2022 Elsevier LtdThe Kaiser Effect is amongst the phenomena that can be detected through the so-called acoustic emission method. This effect can be used to evaluate in-situ stress of the rock. Identification of the point at which the Kaiser Effect is applied is problematic in particular cases where the acoustic parameters increase gradually, and this is where signal analysis methods come into play. In this research, Brazilian test was performed on Phyllite samples and acoustic data was recorded simultaneously. Following an energy-based and count-based parametric approach, the occurrence time and the Kaiser Effect stress were obtained for the test specimens. Then, wavelet transform method was used to estimate the occurrence time of the Kaiser Effect. Continuous wavelet transform analysis was undertaken to analyze the signals. The signal analysis was performed by considering the peak frequency as desired parameter and db5 wavelet as the mother wavelet. In continuous wavelet analysis method, the results were classified into five cluster using the K-means clustering algorithm. Considering the mechanism of the Kaiser Effect, the fifth category was considered for further investigations. The results showed that the Phyllite specimens exhibited good capabilities for recovering the stress memory and evaluating the Kaiser Effect-driven stress, the results further showed a good agreement between the occurrence times obtained from the parametric method and those resulted from continuous wavelet transform technique, so that the corresponding differences in the Kaiser Effect-driven stress are negligible. The level of Kaiser Effect-driven stress obtained from the proposed method were acceptably in agreement with the preloading stress levels.
查看更多>>摘要：? 2022 Elsevier LtdRolling bearing fault diagnosis with imbalanced data is a challenging task. It is a significant means to augment the data into balanced datasets. A novel data augmentation method named CVAEGAN-SM is proposed to address this issue in this paper. Firstly, to alleviate the overfitting of generative models due to data scarcity, the input data is preprocessed with a joint translating and scaling, whose hyperparameters are fed by the self-modulation output parameters. Secondly, concerning the conditional generative adversarial network, self-modulation is embedded into the generator, which allows the generator to update itself simultaneously relying on the feedback of input and discriminator. Thirdly, A novel model is constructed integrating the conditional variational autoencoder and conditional Wasserstein generative adversarial network with self-modulation. Furthermore, multi-class comparative experiments are conducted to demonstrate the effectiveness and performance of CVAEGAN-SM. Experimental results indicate that CVAEGAN-SM can effectively augment the imbalanced dataset and outperforms other well-advanced methods.
查看更多>>摘要：? 2022 Elsevier LtdDeep learning has made remarkable achievements in fault diagnosis. However, the working conditions of the axial piston pump are diverse, and the distribution of the data is not the same, which causes most of the deep learning models to invalid. A deep multi-signal fusion adversarial model based transfer learning (MFAN) is presented to solve this problem. A multi-signal fusion module is designed to assigns weights to vibration signals and acoustic signals, which improves the dynamic adjustment ability of the method. Moreover, the residual network is embedded in the shared feature generation module to obtain abundant feature information. According to the different working loads of the axial piston pump, nine transfer scenarios are designed, and the proposed method is compared with five typical diagnosis methods. The average accuracy of MFAN on all scenarios reaches 98.5%, indicating this method has excellent performance in cross-domain fault detection of axial piston pumps.
查看更多>>摘要：? 2022 Elsevier LtdA complete measurement method dedicated to capacitive sensors has been developed. It includes the development of hardware (an analogue interface circuit for microcontrollers with built-in times/counters and analogue comparators) and software (a measurement procedure and a systematic error calibration (correction) algorithm which is based on a calibration dictionary). The interface circuit consists of a low-pass filter and a phase shifter with a capacitive sensor. A prototype circuit based on a mid-range 8-bit microcontroller has been developed and investigated. We also analysed the relative inaccuracy of a measured capacitance of the sensor and performed experimental research. We obtained the relative errors of capacitance determination < 0.06%, which gives a measurement accuracy < 72 fF for the assumed range of the capacitance (100–300 pF).
查看更多>>摘要：? 2022 Elsevier LtdTo achieve more accurate identification of mechanical faults for high voltage circuit breaker (HVCB) with higher speed, multi-sensor information fusion has been proposed in this paper. The wavelet packet decomposition is used to decompose the signals collected by various sensors. Then, the energy of the wavelet packets in different frequency band can be obtained to constitute eigenvectors. After that, Dempster/Shafer (D-S) evidence theory can be applied for fault identification, where neural networks are built to train the weight of sensors without prior information. The conflicts existing in information fusion can be solved and the accuracy of fault diagnosis is improved. Finally, experimental validation is carried out to show the effectiveness of the proposed fault diagnosis strategy for the HVCB.
查看更多>>摘要：? 2022 Elsevier LtdValves are widely used in industrial and domestic pipeline systems. However, during their operation, they may suffer from the occurrence of the cavitation, which can cause loud noise, vibration and damage to the internal components of the valve. Therefore, monitoring the flow status inside valves is significantly beneficial to prevent the additional cost induced by cavitation. In this paper, a novel acoustic signal cavitation detection framework – based on XGBoost with adaptive selection feature engineering – is proposed. Firstly, a data augmentation method with non-overlapping sliding window (NOSW) is developed to solve small-sample problem involved in this study. Then, the each segmented piece of time-domain acoustic signal is transformed by fast Fourier transform (FFT) and its statistical features are extracted to be the input to the adaptive selection feature engineering (ASFE) procedure, where the adaptive feature aggregation and feature crosses are performed. Finally, with the selected features the XGBoost algorithm is trained for cavitation detection and tested on valve acoustic signal data provided by Samson AG (Frankfurt). Our method has achieved state-of-the-art results. The prediction performance on the binary classification (cavitation and no-cavitation) and the four-class classification (cavitation choked flow, constant cavitation, incipient cavitation and no-cavitation) are satisfactory and outperform the traditional XGBoost by 4.67% and 11.11% increase of the accuracy.
查看更多>>摘要：? 2022 Elsevier LtdHigh-accuracy detection instrument is required to monitor temporal and spatial distribution of methane (CH4) and carbon dioxide (CO2) near fault zones. An environmentally adaptive CH4 and CO2 sensor is developed to achieve low power detection in the wild. A state equivalence correction (SEC) method based on state observer theory is proposed to monitor sensor state and improve measurement accuracy. SEC reduces the measurement error by one order of magnitude. SEC errors are less than 3.0‰ for both CH4 and CO2 in different environments. Geochemical application is conducted in Fengman Seismic Station to monitor CH4 and CO2. The measurement accuracies are better than 99.79% for CH4 and 99.77% for CO2. The measurement precisions are 99.83% for CH4 and 99.82% for CO2. Results validated the sensor presents potential in the monitoring of temporal and spatial distribution of CH4 and CO2.
NSTL
Elsevier