Multiresolution Decomposition Research Articles

Research on bearing fault diagnosis based on novel MRSVD-CWT and improved CNN-LSTM

As a critical component in mechanical equipment, rolling bearings play a vital role in industrial production. Effective bearing fault diagnosis provides a more reliable guarantee for the safe operation of the industrial output. Traditional data-driven bearing fault diagnosis methods often have problems such as insufficient fault feature extraction and poor model generalization capabilities, resulting in reduced diagnostic accuracy. To solve these problems and significantly improve the diagnosis accuracy, this paper proposes a novel fault diagnosis method based on multi-resolution singular value decomposition (MRSVD), continuous wavelet transform (CWT), improved convolutional neural network (CNN) enhanced by convolutional block attention module, and long short-term memory (LSTM). Through MRSVD, the vibration signal is decomposed layer by layer into multiple denoised signals, thus signal noise can be eliminated to the greatest extent to gain the optimal denoised signals; then through CWT, the optimal denoised signals are converted into two-dimensional time-frequency images so that the local and global characteristic information can be fully captured. Finally, through improved CNN-LSTM, feature extraction is greatly enhanced, resulting in high accuracy of fault diagnosis. Lots of experiments are organized to test the performance, and the experimental results show that the proposed method on various datasets has better diagnosis accuracy and generalization ability under different working conditions than other methods.

Movie Scene Simulation and Rendering Based on Virtual Reality Technology

Movie scene simulation and rendering with virtual reality (VR) technology have emerged as a game-changing method in the film industry, providing filmmakers with unparalleled creative opportunities and immersive storytelling experiences. This article investigates the use of VR technology in movie scene simulation and rendering, with a focus on sophisticated modelling techniques and rendering methods. Filmmakers can effectively control scene complexity and enhance rendering speed by utilizing VR-based modelling approaches such as multi-resolution decomposition and level of detail (LOD) rendering. These modelling techniques enable the development of highly complex and realistic virtual worlds, allowing filmmakers to bring their creative visions to life with unparalleled precision. In addition, VR-based rendering technologies, such as viewpoint-based geometric chunking rendering, improve frame rates and overall scene rendering quality, resulting in a seamless and immersive watching experience for audiences. Using VR technology, filmmakers can push the boundaries of traditional filmmaking, create fascinating cinematic experiences, and reimagine narrative in the digital age. This research explores the revolutionary impact of VR technology on movie scene simulation and rendering, emphasizing its potential to change the way films are created, seen, and appreciated by audiences throughout the world. Finally, they compare and assess the scene effect, LOD detail level, and interactive experience effect before and after simplification. scene design for virtual reality technologies generally scores 5-6 on each of the measures, whereas common technology's scenario design usually gets 3 to 4, showing whether the scenario design of virtual reality technology works properly.

New formulation for predicting total dissolved gas supersaturation in dam reservoir: application of hybrid artificial intelligence models based on multiple signal decomposition

Total dissolved gas (TDG) concentration plays an important role in the control of the aquatic life. Elevated TDG can cause gas-bubble trauma in fish (GBT). Therefore, controlling TDG fluctuation has become of great importance for different disciplines of surface water environmental engineering.. Nowadays, direct estimation of TDG is expensive and time-consuming. Hence, this work proposes a new modelling framework for predicting TDG based on the integration of machine learning (ML) models and multiresolution signal decomposition. The proposed ML models were trained and validated using hourly data obtained from four stations at the United States Geological Survey. The dataset are composed from: (i) water temperature (Tw), (ii) barometric pressure (BP), and (iii) discharge (Q), which were used as the input variables for TDG prediction. The modelling strategy is conducted based on two different steps. First, six singles ML model namely: (i) multilayer perceptron neural network, (ii) Gaussian process regression, (iii) random forest regression, (iv) random vector functional link, (v) adaptive boosting, and (vi) Bootstrap aggregating (Bagging), were developed for predicting TDG using Tw, BP, and Q, and their performances were compared. Second, a new framework was introduced based on the combination of empirical mode decomposition (EMD), the variational mode decomposition (VMD), and the empirical wavelet transform (EWT) preprocessing signal decomposition algorithms with ML models for building new hybrid ML models. Hence, the Tw, BP, and Q signals were decomposed to extract the intrinsic mode functions (IMFs) by using the EMD and VMD methods and the multiresolution analysis (MRA) components by using the EWT method. Then after, the IMFs and MRA components were selected and regraded as new input variables for the ML models and used as an integral part thereof. The single and hybrid prediction models were compared using several statistical metrics namely, root mean square error, mean absolute error, coefficient of determination (R2), and Nash–Sutcliffe efficiency (NSE). The single and hybrid models were trained several times with high number of repetitions, depending on the kind of modeling process. The obtained results using single models gave good agreement between the predicted TDG and the situ measured dataset. Overall, the Bagging model performed better than the other five models with R2 and NSE values of 0.906 and 0.902, respectively. However, the extracted IMFs and MRA components using the EMD, VMD and the EWT have contributed to an improvement of the hybrid models’ performances, for which the R2 and NSE were significantly increased reaching the values of 0.996 and 0.995. Experimental results showed the superiority of hybrid models and more importantly the importance of signal decomposition in improving the predictive accuracy of TDG.Graphical abstract

Design of fault location algorithm based on online distributed travelling wave for HV power cable.

To tackle the challenge of localization failure due to traveling wave dispersion during a mid-line fault in a long high-voltage cable, this study conducts an in-depth analysis of the refraction characteristics of the cumulative three-phase sheath currents at cross commutation points and direct grounding points, facilitated by detailed theoretical derivation. We introduce the novel concept of a 'first traveling wave attenuation index,' which quantifies the peak of the initial faulted traveling wave. We strategically place distributed traveling wave detection devices solely in the first direct grounding box of each cross-interconnected main section, effectively segmenting the line into five distinct zones. These zones are identified based on the unique transient polarities exhibited by the first traveling wave of fault current at each impedance mismatch point along the cable. To overcome the issue of weak traveling wave signals collected at the first end measurement point, we propose an innovative peak detection method for the first wave. This method harnesses the power of empirical wavelet transform (EWT) and multi-resolution singular value decomposition (MRSVD), providing a significant boost in ranging accuracy compared to traditional wavelet methods. Simulation results validate the efficacy of our proposed fault location method, which accurately pinpoints the fault section by contrasting the polarity of the first wave peak of the sheath current detected at each measurement point. Notably, our method ensures safety and convenience as the equipment does not require direct contact with high voltage.

Energy Content Analysis of the Large 2019 Ridgecrest, California Seismic Sequence

The energy content of ground motion records obtained during the 2019 M6.4, M7.1 Ridgecrest, California earthquakes which occurred closely in space is analyzed in this paper. Wavelet transform as a powerful tool, that is used to evaluate the energy content characteristics including cumulative energy profiles and the total energy for different components of records, was obtained at three stations named China Lake Christmas Canyon, China Lake and Tower 2. Due to the fact that the energy-based parameters could be well correlated with damage distribution, this path has been taken into consideration in this research work with the light of the powerful wavelet tool. For this purpose, the discrete wavelet transform is efficiently used for multi-resolution decomposition of acceleration time history of ground motions and a new framework is presented to estimate the energy level of earthquake sequences. Finally, the pseudo-period corresponding to each corresponding level is introduced and the distribution of energy vs. the pseudo-period is discussed. The results of this paper provide a platform to compare the energy levels of different earthquake sequences in different periods. Analysis showed that wide-band energy distribution could be identified at China Lake Christmas Canyon and Tower 2 stations.

A versatile Wavelet-Enhanced CNN-Transformer for improved fluorescence microscopy image restoration

Fluorescence microscopes are indispensable tools for the life science research community. Nevertheless, the presence of optical component limitations, coupled with the maximum photon budget that the specimen can tolerate, inevitably leads to a decline in imaging quality and a lack of useful signals. Therefore, image restoration becomes essential for ensuring high-quality and accurate analyses. This paper presents the Wavelet-Enhanced Convolutional-Transformer (WECT), a novel deep learning technique developed specifically for the purpose of reducing noise in microscopy images and attaining super-resolution. Unlike traditional approaches, WECT integrates wavelet transform and inverse-transform for multi-resolution image decomposition and reconstruction, resulting in an expanded receptive field for the network without compromising information integrity. Subsequently, multiple consecutive parallel CNN-Transformer modules are utilized to collaboratively model local and global dependencies, thus facilitating the extraction of more comprehensive and diversified deep features. In addition, the incorporation of generative adversarial networks (GANs) into WECT enhances its capacity to generate high perceptual quality microscopic images. Extensive experiments have demonstrated that the WECT framework outperforms current state-of-the-art restoration methods on real fluorescence microscopy data under various imaging modalities and conditions, in terms of quantitative and qualitative analysis.

An adaptive real-time modular tidal level prediction mechanism based on EMD and Lipschitz quotients method

Real-time prediction of tidal level is vital for on-the-spot activities such as marine transportation and ocean surveys. Aiming at the complex characteristics of nonlinearity, time-varying dynamics, and uncertainty generated by celestial bodies' movements and influenced by geographical as well as hydrometeorological factors, an adaptive real-time modular tidal level prediction mechanism is proposed based on empirical mode decomposition (EMD) and Lipschitz quotients method. An adaptive modular tidal level prediction mechanism is proposed by combining the harmonic analysis method with a variable structure neural network. The order of time series decomposition and the prediction input model order of the neural network are automatically determined based on EMD and the Lipschitz quotients method, respectively. The adaptability of the prediction mechanism is further enhanced with the network dimension, hidden units’ locations, and connecting parameters of the variable neural network being online adjusted in a sequential learning scheme. While the extraction of harmonic components alleviates the difficulty in prediction, the multi-resolution decomposition of residual series provides further insight into the time-varying tide dynamics caused by environmental disturbances, thus enabling precise predictions for tidal levels. The feasibility and effectiveness of the proposed adaptive modular tidal prediction mechanism are demonstrated based on the real-measured tidal level data.

Improvement of emotion classification performance using multi-resolution variational mode decomposition method

Automated speech emotion recognition (SER) has been gaining popularity among researchers for three decades because of its vast number of applications in the real world. It is helpful to improve the relationship between humans and machines, online marketing and education, customer relations, medical treatment, safe driving, online search, etc. Researchers adopt various methods to improve emotion recognition performance from speech signals, like using different combinations of features (acoustic, non-acoustic, or both), classifiers (machine learning, deep learning, or both). Our study tried to improve emotion classification performance using features based on the multi-resolution variational mode decomposition (MRVMD) method. We first decompose each signal frame into several sub-signals known as modes or intrinsic mode functions (IMFs) using the MRVMD method. Then the proposed features, multi-resolution variational mode mel-frequency cepstral coefficient (MRVMMFCC), multi-resolution variational mode approximate entropy (MRVMAE), and multi-resolution variational mode permutation entropy (MRVMPE), were extracted using the MRVMD-decomposed IMF signals. Finally, different combinations of the proposed features are used to classify the emotion using a deep neural network (DNN) classifier. From the experimental results, we found that combination of the proposed feature (MRVMMFCC + MRVMAE + MRVMPE) performs better than the other combination in recognizing emotion using speech signals. The proposed feature combination with a DNN classifier achieved an emotion classification accuracy of 83.4%, 85.01%, and 90.51% for the SAVEE, EMOVO, and EMO-DB datasets, respectively. We found that the proposed MRVMD method performed better than the state-of-the-art methods.

Reduced-order model and attractor identification for large eddy simulation of squirrel cage fan

A large eddy simulation (LES) of a squirrel cage fan (SCF) provides a precise representation of turbulent flows with different degrees of complexity. This study comprehensively analyzes the coherent structures of turbulent flows in an SCF using an LES, proper orthogonal decomposition (POD), dynamic mode decomposition (DMD), and multi-resolution dynamic mode decomposition (mrDMD). An intelligent reduced-order model is established by integrating hierarchical deep learning and the sparse identification of nonlinear dynamics. The result shows that the evolution of the global DMD modes is attenuated due to the spatial distribution variations of localized high-frequency mrDMD modes, along with the fragmented and non-steady development of modal patterns. Unlike POD, DMD quantifies the quality of the impeller inlet environment and captures the antisymmetric low-dimensional flows associated with the shedding of rotating vortex structures. The interaction strength between stationary and dynamic rotating areas is accurately represented by attractors characterized by petal-like structures. The trajectory of the attractors faithfully maps the antisymmetric structural attributes, quasi-periodic behavior, and gradual attenuation characteristics exhibited by DMD modes. The number of petal-like systems and their temporal oscillations are in good agreement with the number of fan blades and their rotational cycles. This study provides new insight into fan engineering to advance flow control strategies and improve the understanding of the underlying flow mechanisms.

Based on Wavelet and Windowed Multi-Resolution Dynamic Mode Decomposition, Transient Axial Force Analysis of a Centrifugal Pump under Variable Operating Conditions

This study analyzes the transient axial force of a centrifugal pump under variable operating conditions using wavelet analysis and a novel technique called windowed multi-resolution dynamic mode decomposition (wmrDMD). Numerically simulating the sampled time series allows the reconstruction of the impeller’s axial force information, providing validation for this innovative data-driven analysis technique. The comparison between the reconstructed results and the original axial force data demonstrates a remarkable agreement, as all data points exhibit error values below 2.49%. The wmrDMD technique systematically decomposes the impeller’s axial force field into dynamically significant modes across various time scales. Removing the mean flow field in this study resolves the transient motion of the impeller’s axial force, facilitating the identification of positions with high-frequency axial force oscillations and fluctuations in intensity amplitude. The high-frequency axial force of the impeller exhibits stable periodic variations within the operating range of 1.0nr-1.0Qr, whereas the changes are insignificant within the range of 0.4nr-0.4Qr. However, within the operating range of 1.0nr-0.4Qr, both the position and intensity amplitude of the axial force exhibit significant variations without a stable trend. Furthermore, cross-wavelet and wavelet coherence analyses reveal that within the operating range of 0.4nr-0.4Qr, the axial forces on the front and rear cover plates show the strongest correlation at the periodic scale. Within the operating range of 1.0nr-1.0Qr, the next highest correlation is observed, whereas the correlation is lowest within the 1.0nr-0.4Qr operating range.

Research on the scene design of virtual reality technology in film and television animation

Abstract In this paper, firstly, we study the modeling of film and television animation based on virtual reality technology and use quadratic polynomial inverse to control the intensity of light source to establish the lighting model. The image alignment is designed for the scene based on multi-resolution decomposition, and the filtered image layer is obtained by calculating the difference between two layers of Gaussian filtered images, and the Laplace pyramid is established to match the image resolution consistently. Then the scene is chunked and locally rendered, and a smooth transition between geometry and texture is achieved by morphing the geometry next to the texture. Finally, the scene effect, LOD detail level and interactive experience effect before and after simplification are compared and analyzed. The scene design of virtual reality technology basically scores 5-6 on each index, while the scene design of common technology only scores 3-4, indicating that the scene design of virtual reality technology works well.

An optimal sensor placement scheme for wind flow and pressure field monitoring

Sensing of wind pressures on wind-sensitive structures and wind speeds in densely populated cities is crucial to ensure structure safety and monitoring urban wind environment. However, these wind pressure or wind speed sensors are often sparsely distributed due to their high cost, and consequently the collected wind pressure or speed information is limited. Under these circumstances, it is particularly important to place these limited number of sensors in the optimal locations with the capacity of extracting the most information. In this study, an optimal sensor placement scheme based on multi-resolution dynamic mode decomposition (mrDMD), a space post-processing function: signed distance function (SDF) and particle swarm optimization-random forest (PSO-RF) is proposed, called mrDMD-SDF-PSO-RF (mrDSPR). The effectiveness of the proposed scheme is verified using two case studies of reconstruction based on sparse sensors: wind flow field of a city block and wind pressure field of a single building. The result indicates that the proposed scheme significantly improves the efficiency of the sensors. Therefore, based on the proposed optimal sensor placement scheme, the locations of placing sensors can be well-designed for various wind engineering problems, which builds a solid foundation of super-resolution reconstruction of wind pressure on buildings and wind flow field in urban environment.

The Performance of a Time-Varying Filter Time Under Stable Conditions over Mountainous Terrain

Eddy-covariance data from five stations in the Inn Valley, Austria, are analyzed for stable conditions to determine the gap scale that separates turbulent from large-scale, non-turbulent motions. The gap scale is identified from (co)spectra calculated from different variables using both Fourier analysis and multi-resolution flux decomposition. A correlation is found between the gap scale and the mean wind speed and stability parameter z/L that is used to determine a time-varying filter time, whose performance in separating turbulent and non-turbulent motions is compared to the performance of constant filter times between 0.5 and 30 min. The impact of applying different filter times on the turbulence statistics depends on the parameter and location, with a comparatively smaller impact on the variance of the vertical wind component than on the horizontal components and the turbulent fluxes. Results indicate that a time-varying filter time based on a multi-variable fit taking both mean wind speed and stability into account and a constant filter time of 2–3 min perform best in that they remove most of the non-turbulent motions while at the same time capturing most of the turbulence. For the studied sites and conditions, a time-varying filter time does not outperform a well chosen constant filter time because of relatively small variations in the filter time predicted by the correlation with mean flow parameters.

Information fusion for fully automated segmentation of head and neck tumors from PET and CT images.

PET/CT images combining anatomic and metabolic data provide complementary information that can improve clinical task performance. PET image segmentation algorithms exploiting the multi-modal information available are still lacking. Our study aimed to assess the performance of PET and CT image fusion for gross tumor volume (GTV) segmentations of head and neck cancers (HNCs) utilizing conventional, deep learning (DL), and output-level voting-based fusions. The current study is based on a total of 328 histologically confirmed HNCs from six different centers. The images were automatically cropped to a 200 × 200 head and neck region box, and CT and PET images were normalized for further processing. Eighteen conventional image-level fusions were implemented. In addition, a modified U2-Net architecture as DL fusion model baseline was used. Three different input, layer, and decision-level information fusions were used. Simultaneous truth and performance level estimation (STAPLE) and majority voting to merge different segmentation outputs (from PET and image-level and network-level fusions), that is, output-level information fusion (voting-based fusions) were employed. Different networks were trained in a 2D manner with a batch size of 64. Twenty percent of the dataset with stratification concerning the centers (20% in each center) were used for final result reporting. Different standard segmentation metrics and conventional PET metrics, such as SUV, were calculated. In single modalities, PET had a reasonable performance with a Dice score of 0.77±0.09, while CT did not perform acceptably and reached a Dice score of only 0.38±0.22. Conventional fusion algorithms obtained a Dice score range of [0.76-0.81] with guided-filter-based context enhancement (GFCE) at the low-end, and anisotropic diffusion and Karhunen-Loeve transform fusion (ADF), multi-resolution singular value decomposition (MSVD), and multi-level image decomposition based on latent low-rank representation (MDLatLRR) at the high-end. All DL fusion models achieved Dice scores of 0.80. Output-level voting-based models outperformed all other models, achieving superior results with a Dice score of 0.84 for Majority_ImgFus, Majority_All, and Majority_Fast. A mean error of almost zero was achieved for all fusions using SUVpeak , SUVmean and SUVmedian . PET/CT information fusion adds significant value to segmentation tasks, considerably outperforming PET-only and CT-only methods. In addition, both conventional image-level and DL fusions achieve competitive results. Meanwhile, output-level voting-based fusion using majority voting of several algorithms results in statistically significant improvements in the segmentation of HNC.

Multiresolution Dual-Polynomial Decomposition Approach for Optimized Characterization of Motor Intent in Myoelectric Control Systems.

Surface electromyogram (sEMG) is arguably the most sought-after physiological signal with a broad spectrum of biomedical applications, especially in miniaturized rehabilitation robots such as multifunctional prostheses. The widespread use of sEMG to drive pattern recognition (PR)-based control schemes is primarily due to its rich motor information content and non-invasiveness. Moreover, sEMG recordings exhibit non-linear and non-uniformity properties with inevitable interferences that distort intrinsic characteristics of the signal, precluding existing signal processing methods from yielding requisite motor control information. Therefore, we propose a multiresolution decomposition driven by dual-polynomial interpolation (MRDPI) technique for adequate denoising and reconstruction of multi-class EMG signals to guarantee the dual-advantage of enhanced signal quality and motor information preservation. Parameters for optimal MRDPI configuration were constructed across combinations of thresholding estimation schemes and signal resolution levels using EMG datasets of amputees who performed up to 22 predefined upper-limb motions acquired in-house and from the public NinaPro database. Experimental results showed that the proposed method yielded signals that led to consistent and significantly better decoding performance for all metrics compared to existing methods across features, classifiers, and datasets, offering a potential solution for practical deployment of intuitive EMG-PR-based control schemes for multifunctional prostheses and other miniaturized rehabilitation robotic systems that utilize myoelectric signals as control inputs.

A robust and secured fusion based hybrid medical image watermarking approach using RDWT-DWT-MSVD with Hyperchaotic system-Fibonacci Q Matrix encryption.

Digital image watermarking, the process of marking a host image with a watermark, is generally used to authenticate the data. In the medical field, it is of utmost importance to verify the authenticity of the data using Medical Image Watermarking (MIW), especially in e-healthcare applications. Recently, MIW with image fusion, the merging of multimodal images to improve image quality, is being widely utilized to make diagnosis more accessible and precise with the verified data. This paper offers a durable and secure fusion-based hybrid MIW approach. The method initially used Fast Filtering (FF) to merge two medical images from different modalities to form the cover image. A first-level Redundant Discrete Wavelet Transform (RDWT) is employed on this host image to locate the component with the highest entropy. Then a single-level Discrete Wavelet Transform (DWT) is applied to it. It performed a Multi-resolution Singular Value Decomposition (MSVD) on the wavelet decomposed component and the embedding watermark. Finally, a Hyperchaotic System-Fibonacci Q Matrix (HFQM) encryption system was utilized, which increases the watermarked image's security. Here, using various medical images, the performance of the proposed technique is evaluated. Without any attacks, the approach achieved a maximum Peak Signal to Noise Ratio (PSNR) of 90.31 dB and a Structural Similarity Index Matrix (SSIM) of value 1. Various watermarking assaults were employed to test the proposed method's resilience. The suggested technique achieved a perfect value of 1 for the Normalised Correlation (NC) for almost all attacks with acceptable imperceptibility, which substantially improves over current procedures. The suggested technique's average embedding and extraction times are 0.3958 and 0.4721 seconds, respectively, which are pretty short compared to existing approaches.

Influence of operating conditions on flow field dynamics and soot formation in an aero-engine model combustor

Large-eddy simulations of a sooting aero-engine model combustor are performed for three operating points. They are analyzed to investigate the influence of secondary air injection and primary equivalence ratio on soot formation dynamics. It will be shown how the causes of a strongly intermittent sooting behavior can be investigated quantitatively by mode analysis and correlated statistics. Simulations rely on Arrhenius reaction rate-based finite-rate chemistry, where 79 species transport equations are solved simultaneously to accurately predict the combustion of ethylene: 43 species represent the gas phase, 36 the soot and soot precursors. Soot evolution is described by a well validated sectional soot model including a sectional model for polycyclic aromatic hydrocarbons and their radicals. This approach captures the influence of changing operating conditions on soot formation well. Hence, soot prediction is found to be accurate enough to investigate the formation processes. The presence of coherent and incoherent flow field dynamics in the combustor necessitates the use of multiresolution proper orthogonal decomposition and correlated statistics for the analysis. While the precessing vortex core is found to support soot production continuously, low frequency dynamics are identified to cause the intermittence at all operating conditions. Secondary air injection significantly increases the intensity of the low frequency dynamics. Yet, the soot volume in the combustor varies by the same order of magnitude without secondary air injection. The soot formation intermittence is closely linked to an axially symmetric mixture fraction mode near the primary injector at all operating conditions. Low mode energy required an additional mode analysis of joint statistics which confirmed this conclusion. Increasing the equivalence ratio at the primary injector decreases the influence of the low frequency dynamics on soot formation.

Wavelet-like denoising of GNSS data through machine learning. Application to the time series of the Campi Flegrei volcanic area (Southern Italy)

The great potential of the Global Navigation Satellite System (GNSS) in monitoring ground deformation is widely recognized. As with other geophysical data, GNSS time series can be significantly noisy, hiding elusive ground deformation signals. Several denoising techniques have been proposed to improve the signal-to-noise ratio over the years. One of the most effective denoising techniques has been proved to be multi-resolution decomposition through the discrete wavelet transform. However, wavelet analysis requires long data sets to be effective, as well as long computation times, that hinder its use as a real or near real-time monitoring tool. We propose training by a Convolutional Neural Network (CNN) to perform the equivalent of wavelet analysis to overcome these limitations. Once trained, the CNN model provides answers within seconds, making it feasible as a real-time data analysis tool. Our Machine Learning algorithm is tested on daily GNSS time series collected in the Campi Flegrei caldera (Southern Italy), which is a highly volcanic risk area. Without significant gaps, the retrieved RMSE and R2 values vary in the ranges 0.65–0.98 and 0.06–0.52 cm, respectively. These results are encouraging, as they hint at the possibility of applying this methodology in more effective real-time monitoring solutions for active volcanoes.

Improving the accuracy of the PET/MRI tridimensional multimodal rigid image registration based on the FATEMD

The subject matter of the article is the improvement in the accuracy of multimodal image registration between PET and MRI images in the medical field. The focus of the article pertains to the importance of these images in diagnosis, interpretation, and surgical intervention. This study increased the accuracy of PET/MRI multimodal image registration achieved through a new approach based on the multi-resolution image decomposition. The tasks to be solved are: The study proposes a new method, the fast and adaptive three-dimensional mode decomposition (FATEMD), to generate multi-resolution components for accurate registration. The method used: The study uses the FATEMD approach, which estimates the transformation parameters of the registration from the PET image and the residue of the second level of the MRI image that is obtained after the extraction of the first two tridimensional intrinsic mode functions (TIMFs). The following results were obtained: The proposed method of multimodal registration between PET and MRI images involves the use of the fast and adaptive three-dimensional mode decomposition (FATEMD) approach. This approach was tested on 25 pairs of images from the Vanderbilt database and was found to have improved accuracy compared to the usual method, as shown through comparative studies using measures of mutual information, normalized mutual information, and entropy correlation coefficient. Conclusion. The main objective achieved in the study was to enhance the accuracy of PET/MRI multimodal image registration through the application of the FATEMD decomposition method. This approach is novel compared to traditional methods as it involves estimating the transformation parameters from the PET image and the second level residue of the MRI image, resulting in more precise outcomes as opposed to using just the PET and MRI images alone. The integration of multiple imaging techniques, such as PET and MRI, provides healthcare professionals with a more comprehensive view of a patient's anatomy and physiology, leading to enhanced diagnosis and treatment planning.

Multi-resolution dynamic mode decomposition for damage detection in wind turbine gearboxes

Abstract We introduce an approach for damage detection in gearboxes based on the analysis of sensor data with the multi-resolution dynamic mode decomposition (mrDMD). The application focus is the condition monitoring of wind turbine gearboxes under varying load conditions, in particular irregular and stochastic wind fluctuations. We analyze data stemming from a simulated vibration response of a simple nonlinear gearbox model in a healthy and damaged scenario and under different wind conditions. With mrDMD applied on time-delay snapshots of the sensor data, we can extract components in these vibration signals that highlight features related to damage and enable its identification. A comparison with Fourier analysis, time synchronous averaging, and empirical mode decomposition shows the advantages of the proposed mrDMD-based data analysis approach for damage detection.