Anti-noise Ability Research Articles

Multipath time delay estimation based on the improved Wigner-ville distribution

The accuracy of underwater multipath channel time delay is crucial in underwater acoustic signal processing. The temporal resolution of the traditional matched filter is restricted by the transmitted signal bandwidth. Wigner-Ville distribution (WVD) which has excellent energy concentration characteristics and effectively avoids the interaction between temporal and frequency resolutions, and is widely used in non-stationary signal processing. However, considering the multipath signal processing, the cross-term in echo WVD would cause inaccurate time delay estimation. Therefore, in this paper, a novel multipath signal time delay estimation method based on the improved WVD is proposed. By utilizing the difference in the continuity of energy changes between auto-term and cross-term in the time-frequency domain, the Seeded Region Growing algorithm is applied and improved to extract auto-term in the WVD. The simulation results show that the method removes cross-term and maintains high concentration characteristics of auto-term, compared with pseudo-WVD and smooth pseudo-WVD, etc. In addition, it improves the accuracy of multipath signal delay estimation and has some anti-noise ability.

Interpretable Multiattribute Predictive Analysis Model Based on Rough Fuzzy Sets and Logistic Regression

The interpretability of prediction models is very important for decision management. The vast majority of existing prediction analysis models based on rough set theory lack a certain interpretability. To break through the existing framework, this paper deeply integrates rough fuzzy sets and logistic regression to construct an interpretable prediction model for multi-attribute information systems. Firstly, the Jensen-Shannon divergence with statistical interpretability is used to capture the attribute information with strong correlation under on a level, and then an fresh attribute-oriented <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\delta$</tex-math></inline-formula> -rough fuzzy set model is presented. The pessimistic and optimistic decision concepts of the proposed rough fuzzy set model are used to to interpretability enhancement of data information. On this basis, an interpretable predictive analysis model is constructed by combining logistic regression model. The construction process of the prediction model is supported by sufficient interpretable information. The predicted result is a result of development trend with a certain interpretability. Finally, to assess the effectiveness and viability of the proposed prediction model, we conduct comparative experiments with other three prediction models accompanied with four prediction performance indexes. Experimental results display that the proposed prediction model has well-pleasing prediction performance and anti-noise ability.

Rotating blade vibration parameter identification based on genetic algorithm

In this paper, aiming at the identification of blade vibration parameters such as constant speed synchronization, constant speed asynchronous, and variable speed synchronization, the simulation platform Simulink is used to model the blade vibration system, and the blade tip timing vibration measurement system model is constructed. A method of blade vibration parameter identification based on genetic algorithm is proposed, and numerical simulation and experimental verification are carried out. The results show that the parameter identification of blade vibration by genetic algorithm has high accuracy and strong anti-noise interference ability. The influence of key parameters on the identification of blade vibration parameters is studied. For the constant speed synchronous and constant speed asynchronous vibration of the blade, the angle between the sensors should not be an integral multiple of 2π as far as possible, and the larger DR (Distribution Range) value should be guaranteed. The higher the frequency doubling of blade vibration, the more sensors are needed. For the variable speed synchronous vibration of the blade, the frequency doubling is greater than the influence of the sensor layout on the parameter identification results, but the number of sensors is too small, which will seriously affect the identification accuracy of the frequency doubling. Aiming at the blade vibration test, a blade vibration tester is designed. The blade variable speed synchronous vibration test is carried out by using the strain gauge method and the tip timing method. The measurement results of the strain gauge method are basically consistent with the measurement results based on the genetic algorithm and the tip timing.

Mechanical fault diagnosis of gas-insulated switchgear based on saliency feature of auditory brainstem response under noise background

The mechanical fault of gas-insulated switchgear (GIS) seriously threatens the security of the power grid. Recently, acoustic-based fault diagnosis methods, which have the advantage of non-contact measurement, have been applied to the GIS mechanical fault diagnosis, but vulnerable to the interference of the background noise. To improve the capacity of the acoustic-based GIS fault diagnosis under noise background, by simulating the sound feature extraction ability and anti-noise ability of human auditory system, a novel GIS mechanical fault diagnosis method based on saliency feature of auditory brainstem response (SFABR) is proposed. First, an auditory saliency model, which considers both the auditory periphery and the auditory nerve center was constructed by combining the deep auditory model and the saliency model. After processing GIS emitted acoustic signal, the auditory brainstem response (ABR) was obtained, and the saliency features of the ABR were extracted to obtain the SFABR. Then, the characteristic frequency distribution of the auditory saliency model was adjusted to make it more suitable for the spectral characteristics of the GIS sound signal. Finally, the SFABR was mapped to a two-dimensional CNN to train a model for GIS mechanical fault diagnosis. This method simulates the process of auditory response extraction and the selection effect of auditory attention on sound elements. The 110 kV three-phase GIS fault simulation experiment shows that for GIS mechanical faults, the diagnosis method based on SFABR can obtain 96.1% fault identification accuracy. In different noise environments, compared with the traditional acoustic-based fault diagnosis methods, this method has stronger anti-noise performance, and can more effectively realize the identification of GIS mechanical failure types. In future research, the method can be further extended to fault diagnosis of more types of power equipment.

Method for Dynamic Load Location Identification Based on FRF Decomposition

The location identification of dynamic load is an important part of load-identification technology. Traditional methods are mostly aimed at the identification of dynamic load’s amplitude and phase. A new method for dynamic load location identification is proposed in this paper. An amplitude ratio or a phase difference between structural dynamic response signals is only related to the frequency response function (FRF), which is a complex-valued function that implies the location information of the load to be identified. In this method, the amplitude and phase variables of the excitation can be eliminated with symbolic calculations, and the location variables can be extracted from the FRF in a complex-number field. An excitation location can be identified quickly with parameter optimization using a Genetic Algorithm (GA), avoiding the ill-posed problem caused by matrix inversion. Numerical simulations and experiments demonstrate that this method can realize the fast recognition of several excitation positions, and has a high recognition accuracy, a short calculation time and a strong anti-noise ability.

Color ghost imaging based on optimized random speckles and truncated singular value decomposition

The ambient noise always harms the image quality in ghost imaging. In this letter, we propose a scheme of color ghost imaging to reduce the noise impact. In addition to the method of truncated singular value decomposition (TSVD), the measurement matrix is optimized with low-pass filters. In experiment, four (Gaussian, averaging, circular averaging, and median) filters are applied to the patterns of random speckles, which reforms the measurement matrix. By using the TSVD method, the pseudo-inverse matrix of the optimized measurement matrix is obtained and used to reconstruct the images. With proper truncation rates, the mean square errors of the reconstructed images are greatly decreased. Therefore, the image quality in our scheme is greatly improved. Further simulation analysis shows that the point spread functions are greatly optimized after filtering the random speckles. Our technique strengthens the anti-noise ability of ghost imaging, especially in complex environment.

Multi-branch convolutional attention network for multi-sensor feature fusion in intelligent fault diagnosis of rotating machinery

Multi-sensor data fusion approaches based on deep learning are widely used for fault diagnosis of rotating machinery. Massive sensor data bring not only abundant information but also technical challenges for the fault diagnosis. The main challenge is how to discriminate the discrepancies between multi-sensor data and efficiently fuze these sensor data to improve diagnostic performance. To overcome the challenge, a novel multi-branch convolutional attention network (MBCAN) is proposed for fault diagnosis of the rotating machinery. In this method, a feature extractor with attention mechanism is established to extract fault-related features from the sensor data and reduce irrelevant noise interference hidden in the sensor data. Meanwhile, a multi-branch convolutional pathway is designed to enrich the features. Furthermore, a feature fusion module is constructed by defining an adaptive weighted fusion rule to fuze the extracted features. The performance of the MBCAN is verified on gearbox and centrifugal pump under different noise environments. The experimental results show that the proposed MBCAN has more excellent anti-noise ability and reliable diagnostic performance than other existing approaches.

A Random Forest Algorithm Combined with Bayesian Optimization for Atmospheric Duct Estimation

Inversion of atmospheric ducts is of great importance in the field of performance evaluation for radar and communication systems. Since the model parameters in machine learning play a crucial role in prediction performance, this paper develops a random forest (RF) model integrated with Bayesian optimization (BO) called BO-RF for atmospheric duct prediction, and the BO is adopted to determine appropriate model parameters during the training process. In addition, the K-fold cross-validation (CV) method is also incorporated into the model to obtain the best model partition and overcome the overfitting problem. To test the performance of the proposed model, the results obtained by the BO-RF are compared with other commonly used methods, such as classical RF, extreme gradient boosting (XGBoost) with/without BO, and K-nearest neighbor (KNN) with/without BO. Comparisons demonstrate that BO-RF has the best accuracy and anti-noise ability for the estimation of duct parameters.

A modal estimation method of rotating blade based on compressed sensing and blade tip timing

Structural modal estimation has consistently remained one of the most crucial areas of research in mechanical vibration analysis and signal processing. It is imperative to accurately monitor the vibration signals of rotor blades and their corresponding structural modal parameters. Blade tip timing (BTT) is a promising approach used to measure vibration and monitor the health of blades. However, most existing BTT methods focus on frequency accuracy, neglecting damping, a key physical quantity that represents the strength of the structure. This research focuses on damping modal estimation of blade vibration and proposes a sparse reconstruction algorithm based on a two-dimensional Laplace wavelet family, which addresses the inherent under-sampled problem of BTT technology while enabling estimations of both the structural modal frequency and modal damping of the blades. Firstly, the proposed algorithm is achieved by designing a Laplace wavelet dictionary based on prior information and using a convex optimization objective function with a regularization term. Secondly, Laplace wavelet dictionary matches the signal better than the traditional Fourier dictionary based on the similarity between the damped vibration signal and the Laplace wavelet, then a sparser representation vector can be obtained. Moreover, the research object is not limited to a single blade, but can be easily expanded to multiple stages and multiple rotating blades monitored simultaneously. Finally, the simulation and physical test results indicate that the proposed method exhibits high reconstruction accuracy, reliability, and anti-noise abilities.

Analysis of the geothermal formation mechanisms in the Gonghe Basin based on thermal parameter inversion

Gonghe Basin is the most critical distribution area of hot dry rock (HDR) geothermal resources in China. Geophysical results (resistivity, velocity, density, etc.) provide an indirect way to image the spatial distribution to understand its genesis, which has a multisolution problem. However, heat flow and thermal conductivity are the most direct and reliable evidence to reveal the distribution of geothermal resources. The conventional thermal conductivity and heat flow obtained based on geologic surveys and rock samples are highly accurate, but it is challenging to characterize the regional distribution characteristics. In this paper, we develop the heat transfer adaptive finite-element equation parameter inversion workflow to invert the thermal parameters and analyze the geothermal formation mechanisms in the Gonghe Basin. First, the synthetic model test verifies that our inversion process has reliable accuracy and strong antinoise ability. For the Gonghe Basin HDR geothermal example, we use high-precision aeromagnetic data to estimate the Curie depth using the improved Parker-Oldenburg method. Then, the regional conductivity thermal and temperature field can be established according to the empirical formula as the initial input model for the inversion procedure. In addition, we collect the temperature data of six HDR geothermal wells across the Gonghe-Guide Basin as a priori constraint information. Combined with the inverted thermal parameters (thermal conductivity, geothermal gradient, and heat flow), geologic structure, and geophysical result, we infer that the high thermal conductivity layer at a depth of 5–10 km provides the main deep heat source channel and the shallow low thermal conductivity fine sandstone layer of Neogene provides good caprock protection for forming a high-temperature geothermal reservoir.

Robust Open-Circuit Fault Diagnosis for PMSM Drives Using Wavelet Convolutional Neural Network With Small Samples of Normalized Current Vector Trajectory Graph

Open-circuit fault is one of the most common faults in permanent magnet synchronous machine (PMSM) drives. The open-circuit fault can cause the obvious change of stator currents of PMSM. Hence, the previous artificial intelligence based-fault diagnosis method mainly relies on the samples extracted from stator currents. However, the large sets of the samples are required due to the variation of the PMSM operating point, increasing the complexity of fault diagnosis. What's more, stator currents are easily affected by the noise, decreasing the accuracy of fault diagnosis. To solve the issues, this paper proposes a robust open-circuit fault diagnosis method using wavelet convolutional neural network with small samples of normalized current vector trajectory graph. The proposed method uses current normalization to establish small sample sets and combines convolutional neural network with discrete wavelet transform to enhance robustness to noise. The proposed fault diagnosis method is validated by simulation and experiment. Both the results show that the proposed method can effectively diagnose 22 kinds of open-circuit fault types (including healthy mode), being with great antinoise ability and robustness to different working conditions.

Robust Fisher-Regularized Twin Extreme Learning Machine with Capped L1-Norm for Classification

Twin extreme learning machine (TELM) is a classical and high-efficiency classifier. However, it neglects the statistical knowledge hidden inside the data. In this paper, in order to make full use of statistical information from sample data, we first come up with a Fisher-regularized twin extreme learning machine (FTELM) by applying Fisher regularization into TELM learning framework. This strategy not only inherits the advantages of TELM, but also minimizes the within-class divergence of samples. Further, in an effort to further boost the anti-noise ability of FTELM method, we propose a new capped L1-norm FTELM (CL1-FTELM) by introducing capped L1-norm in FTELM to dwindle the influence of abnormal points, and CL1-FTELM improves the robust performance of our FTELM. Then, for the proposed FTELM method, we utilize an efficient successive overrelaxation algorithm to solve the corresponding optimization problem. For the proposed CL1-FTELM, an iterative method is designed to solve the corresponding optimization based on re-weighted technique. Meanwhile, the convergence and local optimality of CL1-FTELM are proved theoretically. Finally, numerical experiments on manual and UCI datasets show that the proposed methods achieve better classification effects than the state-of-the-art methods in most cases, which demonstrates the effectiveness and stability of the proposed methods.

Feed-forward cascaded stochastic resonance and its application in ship radiated line signature extraction

Extracting ship-radiated line signatures from intense background noise presents a significant challenge in remote passive sonar detection and identification. While stochastic resonance (SR) has shown promise for enhancing signal-to-noise ratio (SNR), cascaded stochastic resonance (CSR) offers a superior extension by gradually transitioning energy from high to low frequencies, resulting in smoother waveforms and more evident signatures. However, CSR relies heavily on the first level and lacks robustness, especially at slightly lower SNR. To overcome these limitations, we propose a feed-forward cascaded stochastic resonance (FCSR) method that leverages complete target signal information in each level and superimposes it with the output from the last level, leading to gradual improvements in SNR with high robustness. The superposition weights are designed as a function of the number of cascaded levels with an increasing trend to optimize the output of the entire cascaded system. Furthermore, a phase alignment strategy was developed to improve the superposition process. Through theoretical analysis, we demonstrate the effectiveness of the proposed FCSR method. Further simulation analyses demonstrates that FCSR outperforms CSR, with a remarkable 18 dB improvement in filtering performance under low SNR conditions, an average anti-noise ability enhancement of over 10 dB, and a robustness improvement exceeding 30% at −30 dB. We also validate the practicality and effectiveness of our proposed method through application verification, exhibiting excellent enhancement performance. This study illuminates the importance of reutilizing complete target signal information and emphasizes the potential of cascaded systems to extract signatures from heavy background noise.

A deep learning-based approach with anti-noise ability for identification of rock microcracks using distributed fibre optic sensing data

Most of the existing deep learning-based crack identification models can achieve high accuracy when being trained and tested using data split from the same dataset with minimal noise, while perform poorly on field monitoring data with certain level of noise. This research developed a hybrid attention convolutional neural network (HACNN) for rock microcrack identification with enhanced anti-noise ability for distributed fibre optic sensing data. A hybrid attention module was designed and placed next to some certain convolutional layers to enhance the nonlinear representation ability of the proposed model. Two training interference strategies, namely small mini-batch training and adding dropout in the first convolutional layer, were employed to interfere with the training of the HACNN to enhance its robustness against noise. A series of experiments are designed based on the properties of the two training interference strategies to optimize the model parameters. Results showed that the optimized HACNN achieved higher accuracy on datasets with different signal-to-noise ratios compared to other machine learning algorithms, including the support vector machine, the multilayer perceptron, and an existing one-dimensional convolutional neural network. This research demonstrates the potential of establishing a robust DL-based model for identification of rock microcracks from noisy distributed fibre sensing optic data, even when training the model only with a smoothed dataset.

Double-suppressed possibilistic fuzzy Gustafson–Kessel clustering algorithm

Possibilistic fuzzy c-means clustering (PFCM) is an unsupervised hybrid clustering algorithm, which can partly inherit the stability of fuzzy c-means clustering (FCM) algorithm and the noise robustness of possibilistic c-means clustering (PCM) algorithm. However, there are still several limitations in the PFCM when clustering complex data with multiple characteristics: sensitivity to strong noise, coincident clustering problem in multi-class clustering, and disability to deal with strong correlations of feature components. Therefore, a double-suppressed possibilistic fuzzy Gustafson–Kessel​ clustering algorithm (DS-PFGK) is presented. Firstly, we design a new objective function with a modified Mahalanobis distance and an improved weight parameter setting method based on the stability of fuzzy memberships and the anti-noise ability of possibilistic memberships. Moreover, combined with the absolute attribute of possibilistic memberships and the covariance matrix in the modified Mahalanobis distance, an ellipsoidal cluster core which can depict the distribution of sample data is generated to divide each cluster into different regions for selecting appropriate learning objects and identifying noisy data adaptively. Then, a “double-suppressed competitive learning” strategy is designed for the selected learning objects to solve the coincident clustering problem in multi-class clustering by suppressing the nonwinner possibilistic memberships, and to reduce the iteration number by suppressing nonwinner fuzzy memberships and rewarding winner fuzzy memberships. Finally, a segmentation algorithm on noisy color images based on the DS-PFGK is proposed based on the generated cluster cores. Experiments on synthetic datasets, real datasets, and color images show the good performance of the proposed methods compared with several classical and state-of-the-art clustering algorithms.

Mathematical Morphology-Based Fault Data Self-Synchronization Method for Differential Protection in Distribution Networks

Conventional overcurrent protection is hard to apply to modern distribution networks with high penetration of distributed generations and flexible operation modes. Current differential protection (CDP) is considered to be a good solution, but the expensive cost of data synchronization equipment limits its application. Fault data self-synchronization (FDSS) methods realize the approximate data synchronization according to the features of measured currents at both terminals, which makes it possible to achieve the low-cost CDP. However, the existing FDSS methods still have shortcomings such as low accuracy or complicated steps. In this paper, the principles and defects of existing FDSS methods are first analyzed, and then an improved FDSS method based on mathematical morphology (MM) is proposed. This method is realized by MM filter and MM-based peak-valley detection algorithm, which has higher synchronization accuracy and anti-noise ability. More importantly, in this method, the reference time required for synchronization can be obtained in the pre-communication process, which is easily implemented. The simulation based on PSCAD-MATLAB verifies the effectiveness of the proposed method under different fault conditions and its improvement compared with the existing method. Finally, the RTDS-based hardware test platform is used to verify the differential relay prototypes developed based on the proposed method.

Building Tilt Scanning Measurement Based on Binary Linear Regression Analysis

Abstract With the development of modern surveying and mapping technology, 3D laser scanners are increasingly used for building dip scanning. Building inclination characteristics are analyzed using three-dimensional point cloud data and binary linear regression. It can easily and intuitively obtain the inclination of the building. This method extracted the building inclination Angle from the point, line, plane, surface, and other angles. Then, a matching method based on two dimensions and surfaces is proposed, with a strong anti-noise ability. An optimal matching algorithm of structural plane and surface based on quadratic linear regression is proposed. Experiments prove the anti-interference ability and accuracy of this method. It is found that the binary linear regression fitting method can effectively identify the noise signal when there is a lot of noise interference in the point cloud data. It has good anti-interference ability.

SKND-TSACNN: A novel time-scale adaptive CNN framework for fault diagnosis of rotating machinery

Recently, deep learning models represented by convolutional neural networks (CNN) have played an increasingly important role in rotating machinery fault diagnosis (FD). However, complex working conditions and uncertain noise interference pose severe challenges to the accuracy and anti-noise abilities of CNN-based FD models. To address these, based on a selective kernel block (SKB), a novel time-scale adaptive CNN (TSACNN) and neural network denoiser (NND) are constructed in this study, and an FD framework with strong generalization and anti-noise abilities is established through their deep fusion. Based on the end-to-end idea, the kernel adaptive capability of the SKB enables the traditional CNN to dynamically extract the time-domain information of different-scales hidden in signals, thus enhancing the FD accuracy under complex working conditions. The NND constructed based on the SKB (named SKND) overcomes the defect of information loss in the process of encoding and decoding and effectively improves the anti-noise ability of the TSACNN by improving the signal-to-noise ratio of the input signals. Experimental results show that TSACNN can achieve better performance than four state-of-the-art CNN-based FD models under both noise interference and complex working conditions and that SKND can achieve better denoising ability than a state-of-the-art NND model. By deep fusion of SKND and TSACNN, the proposed SKND-TSACNN can achieve satisfactory FD accuracy and robustness under complex working conditions and strong noise interference.

Design and Analysis of Two Nonlinear ZNN Models for Matrix LR and QR Factorization With Application to 3-D Moving Target Location

Two nonlinear zeroing neural network (ZNN) models with prescribed-time convergence for time-dependent matrix LR and QR factorization are proposed in this paper. To do so, two algorithms and two error functions are constructed to transform the time-dependent matrix LR and QR factorization problems into time-dependent linear equation systems, respectively. Simultaneously, a new activation function is introduced based on the initial ZNN models for the prescribed-time convergence of models. The excellent performance (robustness and convergence) of the two proposed ZNN models are analyzed theoretically. Furthermore, the prescribed-time convergence and anti-noise abilities of the proposed ZNN models are well demonstrated in numerical experiments. Finally, the proposed ZNN model is applied to the moving target location problem, and the results show that the location error is at the millimeter level.

The Least Squares Time Element Method Based on Wavelet Approximation for Structural Dynamic Load Identification

Dynamic load identification is a commonly used and quite important approach to obtain the excitation loads of structures in engineering practice. In this paper, a novel dynamic load identification method combining the least squares time element method (LSTEM), wavelet scaling function and regularization method is proposed, which performs a better accuracy and a stronger anti-noise ability. It decomposes the time history of dynamic load into a series of time elements, and approximates the load profile at each time element using wavelet scaling functions. In order to balance the accuracy and efficiency for load identification, an optimal wavelet resolution is then determined. Simultaneously, the least squares time element model is derived which establishes the forward model for computing the wavelet coefficient. Finally, the wavelet coefficients for dynamic load identification are accurately and stably solved by implementing regularization. By this method, on the one hand, the wavelet scaling function and LSTEM improve the identification accuracy, and on the other hand, the integral process in the least squares operation gains the anti-noise ability for the load identification. A numerical example of a roof structure and an experiment of a composite laminate are studied and verify the effectiveness of the proposed method.