Dynamic Detection Approach Research Articles

Dynamic detection of offshore wind turbines by spatial machine learning from spaceborne synthetic aperture radar imagery

In order to change the energy supply structure of traditional thermal power generation with excessive carbon emissions, more and more offshore wind turbines are being installed and put into operation worldwide. However, the dynamic monitoring of offshore wind turbines is still an emerging problem to be solved at a global scale. This paper proposes an effective approach for dynamic detection of offshore wind turbines by machine learning from spaceborne synthetic aperture radar (SAR) Sentinel-1 satellite data. First, the radar images are preprocessed to reduce the speckle noise and ocean wave clutter. Specifically, the time series radar data are cumulatively averaged according to the monthly interval to eliminate the influence of moving temporary objects such as ships; then, the speckle noise is suppressed using refined lee filter; the imagery affected by ocean wave clutters are also mitigated by using a constant false alarm rate (CFAR) technique. Next, representative data are carefully selected as the labeled training data set; and the random forest (RF) model is trained on the Google Earth Engine (GEE) cloud computing platform. Validation demonstrates an overall accuracy of 99.99%, a producer accuracy of 100%, and an user accuracy of 94.12%. Then, mathematical morphology-based time series spatial data differentiation is proposed for monitoring the change of wind turbine. Finally, the dynamic change of the offshore wind farms in the Yellow Sea of China and the North Sea of Europe are monitored by using the proposed approach from 2015 to 2021. Cross-comparisons with ground truth data show an accuracy of 93.67% for dynamic detection of offshore wind turbines. From a large number of experimental results, it is shown that the proposed approach has the advantages of large scale monitoring and high precision, and can be used for global dynamic detection of offshore wind turbines.

A Novel Morphological Analysis based Approach for Dynamic Detection of Inflected Gujarati Idioms

A Novel Morphological Analysis based Approach for Dynamic Detection of Inflected Gujarati Idioms

ECHAD: Embedding-Based Change Detection From Multivariate Time Series in Smart Grids

Smart grids are power grids where clients may actively participate in energy production, storage and distribution. Smart grid management raises several challenges, including the possible changes and evolutions in terms of energy consumption and production, that must be taken into account in order to properly regulate the energy distribution. In this context, machine learning methods can be fruitfully adopted to support the analysis and to predict the behavior of smart grids, by exploiting the large amount of streaming data generated by sensor networks. In this article, we propose a novel change detection method, called ECHAD (Embedding-based CHAnge Detection), that leverages embedding techniques, one-class learning, and a dynamic detection approach that incrementally updates the learned model to reflect the new data distribution. Our experiments show that ECHAD achieves optimal performances on synthetic data representing challenging scenarios. Moreover, a qualitative analysis of the results obtained on real data of a real power grid reveals the quality of the change detection of ECHAD. Specifically, a comparison with state-of-the-art approaches shows the ability of ECHAD in identifying additional relevant changes, not detected by competitors, avoiding false positive detections.

Single Molecule FRET - A Multi-Environment Ruler for Determining Structure and Dynamics

Studies of multi-domain proteins and multi-protein machines necessitate a deeper understanding of dynamic structure and transient conformations. Therefore, conventional structure determination methods need to be integrated into dynamic detection approaches. We developed a hybrid method that integrates x-ray structure information into self-consistent distance networks based on single-molecule Förster resonance energy transfer (FRET) [1,2]. By analyzing time-correlated distance distributions globally, we can separate real protein dynamics and fluctuations from modelling uncertainties, and ultimately generate time-correlated structural ensembles. On top of that, we reveal correlated small- and large-scale intra-molecular fluctuations from time-resolved, color- and polarization-sensitive fluorescence measurements. We applied our approach to the heat shock protein Hsp90. The chaperone activates large sets of signal transduction proteins often assisted by co-chaperones [3]. These conformation- and nucleotide-dependent processes are not yet comprehensively understood. Our approach resembled the x-ray structure of Hsp90's closed conformation with an RMSD of 2.8 Å. Beyond that, we resolved the previously unknown dynamic open structure of this multidomain protein. The large-scale fluctuations on the lower millisecond timescale might be the basis of a general regulation mechanism. Finally, I want to show how FRET can be orthogonally integrated into cross-disciplinary platforms such as a microfluidics [4] and others, enabling the simultaneous detection of dynamic structure and low-affinity interactions. [1] Hellenkamp B, Wortmann P, Kandzia F, Zacharias M, Hugel T. Multidomain structure and correlated dynamics determined by self-consistent FRET networks. Nature Methods 14(2), 174-180 (2017) [2] Hellenkamp B, Schmid S et al. Precision and accuracy of single-molecule FRET measurements—a multi-laboratory benchmark study Nature Methods 15, 669-676 (2018) [3] Taipale, M et al. Quantitative Analysis of Hsp90-Client Interactions Reveals Principles of Substrate Recognition. Cell 140(5), 987-1001 (2012) [4] Hellenkamp B, Thurn J, Stadlmeier M, Hugel T., submitted

An approach for dynamic detection of inefficient supercomputer applications

Supercomputers play a very important role in science, since they are used in almost all its spheres. Medicine, chemistry, weather prediction, oil and gas exploration, astronomy – in all these and many more areas, high-performance computing is heavily used. But overall efficiency of supercomputer is usually low due to a number of applications with efficiency issues running on a supercomputer. To solve this problem, the authors are developing a tool that analyzes the flow of applications launched on a supercomputer and identifies programs with suspiciously inefficient behavior. Proposed methods take monitoring data as an input since such data provide full and detailed information about application behavior during its execution.We have already conducted a study on identifying applications with abnormal efficiency in the supercomputer job flow. This classification method based on Random Forest algorithm showed good performance results on real-life data of the Lomonosov-2 supercomputer. But this previously proposed method of analyzing the supercomputer job flow has a big drawback – it is not applicable to the analysis of running jobs.In this work we propose a new method that uses Long short-term memory (LSTM) neural network along with our previous classifier to detect running applications with suspicious behavior. In this method, LSTM detects suspicious parts of the application execution by analyzing its previous behavior during this launch, and after detection, our more precise previous classifier concludes whether the application really has a suspicious behavior. This method is currently being developed and tested on real-life applications of the Lomonosov-2 supercomputer.

DRDDR: a lightweight method to detect data races in Linux kernel

Data races in parallel programs are notoriously difficult to detect and resolve. Existing research has mostly focused on data race detection at the user level and significant progress has been made in this regard. However, a very few efforts have attempted to address OS kernel level races. In contrast to user-level applications, the code in kernels uses vastly more complicated synchronization mechanisms, including different kinds of locks, hardware and software interrupts, widely used signal/wait primitives, and various types of low-level shared resources. It is therefore difficult to apply detection methods designed for user-level applications to identify OS kernel level races. In this paper, we present a new detection tool that is able to effectively detect race conditions in the Linux kernel environment. We use a dynamic detection approach, employing hardware debug registers available on commodity processors, to catch races on the fly during runtime. Preliminary experimental results show that our tool can effectively identify real data race instances.

Dynamic Sensing of Localized Corrosion at the Metal/Solution Interface

A Mach-Zehnder interferometer is employed to detect localized corrosion at the metal/solution interface in the potentiodynamic sweep of the iron electrode in solutions. During the electrochemical reactions, local variations of the electrolyte's refractive index, which correlate with the concentration of dissolved species, change the optical path length (OPL) of the object beam when the beam passes through the electrolyte. The distribution of the OPL difference was obtained to present the concentration change of the metal ions visually, which enable direct evidence of corrosion processes. The OPL difference distribution shows localized and general corrosion during the anodic dissolution of the iron electrode in solutions with and without chloride ions, respectively. This method provides an approach for dynamic detection of localized corrosion at the metal/solution interface.

Feature representation and selection in malicious code detection methods based on static system calls

Currently almost all static methods for detecting malicious code are signature-based, this leads the result that viruses can easily escape detection by simple mechanisms such as code obfuscation. In this paper, a behavior-based detection approach is proposed to address this problem. The behaviors of interest are defined as static system call sequences. Unlike the traditional approach, which derives system call sequences by running executables (i.e., dynamic system call sequences), this approach statically analyzes binary code to derive system call sequences. In this paper, a method for deriving static system call sequences is presented, and two automatic feature-selection methods based on n-grams are proposed. We use machine-learning methods, including the K-nearest neighbor, Support Vector Machine, and decision tree methods to classify executables. The proposed approach is compared with the dynamic detection approach using dynamic system call sequences. The experimental results show that the proposed approach has higher accuracy and a lower false positive rate than the dynamic detection approach.

Modeling damage in rotorcraft flexbeams using wave mechanics

The application of wave propagation analysis to detect damage in rotorcraft flexbeams in the form of transverse cracks and delaminations is discussed. The approach used involves examining the local scattering properties of structural discontinuities in the frequency domain to assess the effect of damage on the modal response of the structure. A model using a simulated cantilevered beam is used to illustrate the performance of the wave model approach for dynamic detection of damage in the form of transverse cracks imparted to composite rotorcraft flexbeams. Experimental validation is carried out on a graphite/epoxy (AS4/3501-6) cantilever beam test specimen with ply orientation in vacuum at various RPM levels. Damage in the form of a transverse crack, which extends across the entire width of the specimen is imparted at depths of 25, 50 and 75% of the beam thickness to the test specimen. A piezoelectric actuator mounted near the root is used to excite the structure dynamically to help locate and determine the extent of damage. The resonant frequencies and scattering properties are used as performance metrics to compare the experimental behavior against the analytical predictions from the wave modeling.

Temporal window system: a new approach for dynamic detection — application to surface acoustic wave gas sensors

The use of artificial neural networks-based pattern recognition techniques is now frequent and efficient in the gas sensor signal processing domain. Neural network data sets are generally built with steadystate sensor responses. Nevertheless, when the detection speed is an essential parameter, they must be monitored in a real time mode. In this paper, we present a new dynamic approach and illustrate it with surface acoustic wave sensor NO 2 responses. A filter and a detector constitute the system. They are both implemented with neural networks and both use shifting temporal windows. The filter is based on a recirculation first neural network whereas a backpropagation second neural network is used for the detector whose output is compared to a threshold to turn on an alarm. The aim is to realize a smart sensor. Since no theoretical results exist yet to find the optimal size, initialization and parametrization of backpropagation neural networks, we have studied the influences of weight initialization, temporal window width, hidden neuron number and learning rule parameters. The results show that the best convergence speeds are obtained for weight initial values that are very dependent on the network topology. It was also found that the learning quality and generalization properties were independent of the weight initialization. Other results on the network architecture and performances are presented and discussed.