•New processing method for low-speed bearing fault diagnosis using vibration signals.•Combines pre-whitening and cross-correlation to highlight bearing fault vibration.•Cross-correlation of vibration ...signal and envelope amplifies bearing fault.•Requires no historic data or experimental settings to function properly.•Its use is illustrated for the bearing fault diagnosis at 20 rpm.
Rolling-element bearings are crucial components in all rotating machinery, and their failure will initially degrade the machine performance, and later cause complete shutdown. The period between an initial crack and complete failure is short due to crack propagation. Therefore, early fault detection is important to avoid unexpected machine shutdown and to aid in maintenance scheduling. Bearing condition monitoring has been applied for several decades to detect incipient faults at an early stage. However, low-speed conditions pose a challenge for bearing fault diagnosis due to low fault impact energy. To reliably detect bearing faults at an early stage, a new method termed Whitened Cross-correlation Spectrum (WCCS) is proposed. The method computes the cross-correlation between the whitened vibration signal and its envelope. In this paper, it is detailed how this correlation can improve the fault diagnosis compared to analyzing the envelope spectrum alone. Compared to other methods reported in the literature, the WCCS provides accurate fault detection without involving experimentally tuned settings or bandpass-filtering. Vibration data at 20 rpm rotational speed from an accelerated life-time test of a 40 mm bore size bearing is used to verify the performance of the proposed method. An additional case study using the WCCS on a difficult dataset from the Case Western Reserve University database is also presented to verify the performance.
This article proposes an active learning scheme to detect multiple faults in permanent magnet synchronous motors in dynamic operations without using historical labelled faulty training data. The ...proposed method combines the self-supervised anomaly detector based on a local outlier factor (LOF and a deep Q-network (DQN) supervised reinforcement learner to classify interturn short-circuit, local demagnetisation, and mixed faults. The first fault, which is detected by LOF and verified by an expert during maintenance, is used as training data for the DQN classifier. From that point onward, the LOF anomaly detector and DQN fault classifiers are working in tandem in the identification of new faults, which require expert intervention when either of them identifies a fault. The robustness of the scheme against dynamic operations, mixed fault, and imbalanced training datasets is validated via a comparative study using stray flux data from an in-house test setup.
Rolling element bearings are crucial components in rotating machinery, and avoiding unexpected breakdowns using fault detection methods is an increased demand in industry today. Variable speed ...conditions render a challenge for vibration-based fault diagnosis due to the non-stationary impact frequency. Computed order tracking transforms the vibration signal from time domain to the shaft-angle domain, allowing order analysis with the envelope spectrum. To enhance fault detection, the bearing resonance frequency region is isolated in the raw signal prior to order tracking. Identification of this region is not trivial but may be estimated using kurtosis-based methods reported in the literature. However, such methods may fail in the presence of relatively strong non-Gaussian noise. Cepstrum pre-whitening has also been proposed for this diagnosis challenge, however the noise floor may increase significantly from the normalization of the entire spectrum. In this paper, a new approach for identifying multiple resonance regions is proposed. The proposed method highlights all resonance frequencies in the signal by combining computed order tracking and cepstrum pre-whitening in a new way. Simulations and experimental results prove the validity of the method, and comparisons with two existing methods show the increase in effectiveness of the proposed method.
This paper aims to classify local demagnetisation and inter-turn short-circuit (ITSC) on position sensorless permanent magnet synchronous motors (PMSM) in transient states based on external stray ...flux and learning classifier. Within the framework, four supervised machine learning tools were tested: ensemble decision tree (EDT), k-nearest neighbours (KNN), support vector machine (SVM), and feedforward neural network (FNN). All algorithms are trained on datasets from one operational profile but tested on other different operation profiles. Their input features or spectrograms are computed from resampled time-series data based on the estimated position of the rotor from one stray flux sensor through an optimisation problem. This eliminates the need for the position sensors, allowing for the fault classification of sensorless PMSM drives using only two external stray flux sensors alone. Both SVM and FNN algorithms could identify a single fault of the magnet defect with an accuracy higher than 95% in transient states. For mixed faults, the FNN-based algorithm could identify ITSC in parallel-strands stator winding and local partial demagnetisation with an accuracy of 87.1%.
In this article, a novel multilevel topology for three-phase applications, having three-level and hybrid N -level modular configurations, enabling low-, medium-, and high-voltage operations, is ...presented. The proposed topology has several attractive features, namely reduced component count, being capacitor-, inductor-, and diode-free, lowering cost, control-complexity, and size, and can operate in a wide range of voltages and powers. Selected simulation and experimental results are presented to verify the performance of the proposed topology. Further, the overall efficiency of the topology and loss distribution in switches are studied. Finally, the key features of the proposed topology in terms of component count, blocking voltage, and dc-link requirements are highlighted via a comparative study.
This paper proposes a novel three-phase topology with a reduced component count for low- and medium-voltage systems. It requires three bidirectional switches and twelve unidirectional switches for ...producing four-level voltages without using flying capacitors or clamping diodes, reducing the size, cost, and losses. Removing flying capacitors and clamping diodes allows it to simplify control algorithms and increase the reliability, efficiency, and lifetime. A modified low-frequency modulation (LFM) scheme is developed and implemented on the proposed topology to produce a staircase voltage with four steps. Further, a level-shifted pulse width modulation (LSPWM) is used to reduce the filter size and increase the output voltage controllability. In this study, a voltage balancing control algorithm is executed to balance the DC-link capacitor voltages. The performance of the proposed topology is numerically demonstrated and experimentally validated on an in-house test setup. Within the framework, the power loss distribution in switches and conversion efficiency of the proposed circuit are studied, and its main features are highlighted through a comparative study.
This paper proposes a novel three-phase transformer-based multilevel inverter (MLI) topology to maximize the output voltage levels for high-power high-voltage applications while reducing component ...counts as compared to its transformer-based counterparts. The proposed hybrid topology is formed by connecting a three-level T-type module with full H-bridge cells through single-phase transformers. The T-type module is fixed while the full H-bridge cell can be repeated for enlarging the output voltage levels without increasing the voltage stress on switches. Key features of the proposed topology include low part count, capacitor-free, diode-free, voltage boosting, simple control, and modularity. Within the framework, a simple low-frequency pulse width modulation (LFPWM) switching scheme is used to control the output voltage, and the working principle is detailed for seven-, nine-, and N-level operation. The operability and performance of the proposed topology are numerically verified and experimentally validated at different loads. Moreover, its conversion efficiency is experimentally examined. Finally, a comparative study with existing transformer-based MLI circuits is conducted to prove its key merits.
Image-based 3D reconstruction is a challenging task that involves inferring the 3D shape of an object or scene from a set of input images. Learning-based methods have gained attention for their ...ability to directly estimate 3D shapes. This review paper focuses on state-of-the-art techniques for 3D reconstruction, including the generation of novel, unseen views. An overview of recent developments in the Gaussian Splatting method is provided, covering input types, model structures, output representations, and training strategies. Unresolved challenges and future directions are also discussed. Given the rapid progress in this domain and the numerous opportunities for enhancing 3D reconstruction methods, a comprehensive examination of algorithms appears essential. Consequently, this study offers a thorough overview of the latest advancements in Gaussian Splatting.
Pitch systems are one of the components with the most frequent failure in wind turbines. This paper presents a two-stage fault detection and classification scheme for electric motor drives in wind ...turbine pitch systems. The presented approach is suitable for application in offshore wind farms with electric pitch systems driven by induction motors as well as permanent magnet synchronous motors. The adopted strategy utilizes three-phase motor current sensing at the pitch drives for fault detection and only when a fault condition is detected at this stage, features extracted from the current signals are transmitted to a support vector machine classifier located centrally to the wind farm. The proposed method is validated in an in-house setup of the pitch drive.
This paper proposes a method for calculating a health indicator (HI) for low-speed axial rolling element bearing (REB) health assessment by utilizing the latent representation obtained by variational ...inference using Variational Autoencoders (VAEs), trained on each speed reference in the dataset. Further, versatility is added by conditioning on the speed, extending the VAE to a conditional VAE (CVAE), thereby incorporating all speeds in a single model. Within the framework, the coefficients of autoregressive (AR) models are used as features. The dimensionality reduction inherent in the proposed method lowers the need of expert knowledge to design good condition indicators. Moreover, the suggested methodology allows for setting the probability of false alarms when encoding new data points to the latent variable space using the trained model. The effectiveness of the proposed method is validated based on two different datasets: from a workshop test of an offshore drilling machine and from an in-house test rig for axial bearings. In both datasets, the HI is exceeding the warning and alarm levels with a probability of false alarm (PFA) of 10 -6 , and the method is most effective at lower shaft speeds.