Panigrahi, Rasmi RanjanMishra, ManoharNayak, JanmenjoyShanmuganathan, Vimal...
15页
查看更多>>摘要：Presently, the issue of power quality (PQ) disturbances in electrical power system has been greater than before owing to increased use of power electronics based nonlinear loads. This work has proposed a hybrid PQ detection and classification algorithm that uses fast-discrete-S-transform (FDST) as feature extraction (FE) technique and memetic firefly algorithm (MFA) based Light-gradient-boost-machine (LGBM) as a classifier. In general, 25 types of PQ signals, comprising both single and multiple disturbances, are studied considering the IEEE-1159 standard. A 3.2 kHz sampling frequency is used on ten cycles of distorted waveforms for the FE. The experimental results clearly proves the effectiveness of the proposed approach with high detection accuracy (99.714% with synthetic data and 99.66% with simulated data), less computational complexity and immune to noisy environments. To end, this work has performed a comparative study with other contemporary FE techniques and classifiers, and in addition with other previously published work.
查看更多>>摘要：Uncalibrated capacitive voltage transformers (CVTs) may significantly degrade measurement accuracy, because of the undetected excessive measurement error (ME). In this article, an online detection method is proposed which combines multi-source heterogeneous data composed of CVT measurements, acceptance test errors, and error limits. By measuring the same voltage with multiple CVTs, the monitoring statistics are generated and the statistic thresholds for the excessive ME detection are set according to the acceptance test errors and the error limits. To further ensure accuracy, the monitoring statistics and acceptance test errors for the CVTs surpassing the thresholds are used to estimate the ME. This estimation is then compared with the error limits as a cross-check to the detection result. Simulation shows that the difference between the ME estimated from the proposed method, and the actual ME is less than 0.01 % and the faulty CVT recognition accuracy exceeds 99%.
查看更多>>摘要：The degradation process of lithium-ion batteries has memory, i.e. it has long-range dependence (LRD). In this paper, an iterative model of the generalized Cauchy (GC) process with LRD characteristics is proposed for the remaining useful life (RUL) prediction of lithium-ion batteries. The GC process uses two independent parameters, fractal dimension and Hurst exponent, to measure the LRD of the degradation process. The diffusion term of the GC iterative model is replaced by the increment of the GC time sequences, constructed via the autocorrelation function (ACF) to describe uncertainty and the LRD characteristics of the lithium-ion batteries capacity degradation. Linear and nonlinear drift terms are used to explain the degradation trend of the lithium-ion batteries capacity. A comparison is made with fractional Brownian motion (FBM) and long-short-term memory (LSTM) network models to show how the GC iterative model has the best performance in RUL prediction of lithium-ion batteries.
查看更多>>摘要：In this work, temperature distribution of aluminium alloy (Al 6061-T6) during the cold forging process is studied with the support of inverse analysis. The values of friction at die-work interface, Taylor-Quinney coefficient and interface heat transfer coefficient (IHTC) are estimated using the inverse method with the help of finite element method (FEM) and experiments on a cylindrical specimen. These values are used to predict the temperature distributions during forging of workpieces of different geometries i.e., hollow cylinder and frustum of cone specimens. Taylor-Quinney coefficient may differ considerably depending on the material and processing condition. Hence, it is recommended to obtain the proper value from an inverse procedure as proposed in this work. Transient analysis of a frustum of cone specimen during the cold forging process is carried out and the temperature contours are presented. A good agreement is seen between the FEM simulation and experimental results.
查看更多>>摘要：The paper considers improving accuracy of measured parameters and quality factors of reactive components. To increase the accuracy of measurement, circuit diagrams of new measuring transducers for measuring the quality factor of reactive elements for operation in a wide range of frequencies are developed. These measuring transducers provide reducing the standard uncertainty of the quality factor measurement for reactive elements with a slight complication of the measuring instrument. A phase and amplitude-phase methods of measuring the quality factor of reactive elements are considered, which provide the measurement standard uncertainty reduction. These methods have a relatively high measurement accuracy and are easy to implement on the microprocessor-based technology. Quality factors of inductive and capacitive elements are shown to be determined by measuring the phase shift angle between two output voltages of the measuring transducer. A prototype of the measuring transducer was installed to verify experimentally obtained theoretical results. Studies were performed for three inductors of different types and for low-power high-frequency capacitors. Theoretical and experimental dependences were investigated and compared.
查看更多>>摘要：A deep learning-based defect identification scheme for the Pelton wheel has been developed. Initially, the raw vibration signal is passed through a time-varying filter based empirical mode decomposition (TVF-EMD). Filter parameters of TVF-EMD are optimized by a newly developed optimization algorithm i.e., amended grey wolf optimization (AGWO) with Kernel estimate for mutual information (KEMI) as its fitness function. The prominent IMF obtained is used to construct scalogram and prepare dataset. The training dataset trains the convolutional neural network (CNN) model whose accuracy was evaluated by the test dataset and founds to be 100%. The proposed AGWO algorithm was evaluated on twenty-three classical benchmark functions and the Wilcoxon test. Results obtained at benchmark functions and the Wilcoxon test validate the efficiency and superiority of the proposed method as compared to other techniques. A CNN classifier is compared with other learning models which suggested that CNN outperforms all learning models.
查看更多>>摘要：Circle and square grid methods are widely used for strain analysis in sheet metal forming operations. In the current paper, images of the deformed grids are captured using a low-cost generic USB microscope and analyzed using an image processing algorithm for automatic strain measurement. The images of the deformed circular grid are analyzed using the segmentation approach, arc support line detection algorithm, and least squared errorellipse fitting algorithm. In case of the deformed square grids, the corners of the quadrilateral are found using Hough Transform (HT) with clustering and are used to fit the maximum size ellipse inside the quadrilateral by projective collineation method. The effectiveness of different algorithms in measurement has been discussed. Good agreement was found between the results obtained from experiments and simulations in LS-Dyna software. The software is able to measure the strain with a maximum error of 1.933 % and 2.11 % for circular and square grids, respectively.
查看更多>>摘要：As for vision-based pose estimation, which is also known as the PnP problem, non-iterative algorithms are more efficient. Precise extraction of 2D projections of feature points is important. If the projections of the feature points are not accurately extracted, the pose estimation accuracy is reduced. Under the condition of natural light, a camera captures the images of feature points, and the existence of high-light regions in the image affects the extraction accuracy of 2D projections of feature points, which reduces the number of effective feature points and leads to poor pose estimation accuracy. In the redundant cases (n > 4), redundant feature points are introduced as additional information, increasing the number of effective feature points to reduce the impact of high-light regions and improve the pose estimation accuracy. For the non-redundant cases (n = 4), it was difficult to ensure pose estimation accuracy. To solve this problem, a non-iterative pose estimation method based on the optimum polarization angle via four corner points of a parallelogram was proposed in this study. First, a model for solving optimum polarization angle was established. Thereafter, on the premise of the optimum polarization angle, the images were captured. Finally, the projections of the four corner points of a parallelogram were extracted, and the object pose was solved non-iteratively according to the four corner points. The corner point extraction experimental results show that the slope difference between the two parallel sides of each parallelogram under the condition of optimum polarization angle is less than that under the condition of natural light, thereby proving the improvement of the imaging quality. Measurement accuracy verification experiments prove that our pose estimation algorithm and the optimum polarization angle is the best combination to improve the non-iterative pose estimation accuracy in non-redundant cases. In the measurement range of - 60-+60 degrees, the angle measurement error was less than +/- 0.0491 degrees. In the measurement range of 0-20 mm, the displacement measurement error was less than +/- 0.0435 mm.
查看更多>>摘要：The aim of this paper is to present an experimental analysis of a solar module emulator based on a low-cost microcontroller. An Arduino low-cost microcontroller is used to mimic several solar module technologies with high accuracy. The estimation and extraction of the single-diode photovoltaic model parameters is performed by a developed software. Two control strategies are applied to the photovoltaic emulation, one based on the digital proportional integral controller and the second one based on fuzzy logic controller. The emulator is able to follow the module I-V curves for fast and simultaneous real-time change of solar radiation, cell temperature and load, based on an online resolution of the single-diode mathematical model of the photovoltaic module. Finally, the experimental and modeled I-V curves of four different photovoltaic module technologies are compared and assessed the efficiency of the proposed fuzzy logic control strategy compared to the proportional integral control strategy.
查看更多>>摘要：In the Engineering discipline, prognostics play an essential role in improving system safety, reliability and enabling predictive maintenance decision-making. Due to the adoption of emerging sensing techniques and big data analytics tools, data-driven prognostic approaches are gaining popularity. This paper aims to deliver an extensive review of recent advances and trends of data-driven machine prognostics, with a focus on their applications in practice. The primary purpose of this review is to categorize existing literature and report the latest research progress and directions to support researchers and practitioners in acquiring a clear comprehension of the subject area. This paper first summarizes fundamental methodologies on data-driven approaches for predictive maintenance. Then, the article further conducts a comprehensive investigation on the different fields of applications of machine prognostics. Finally, a discussion on the challenges, opportunities, and future trends of predictive maintenance is presented to conclude this paper.
NSTL
Elsevier