PL
 |
EN
Preferencje help
Polish version English version Język
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 5

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
Wyszukiwano:
w słowach kluczowych:  data streaming
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
1
Content available An algorithm for arbitrary-order cumulant tensor calculation in a sliding window of data streams
Domino Krzysztof, Gawron Piotr
International Journal of Applied Mathematics and Computer Science
|
2019
|
Vol. 29, no. 1
195--206
EN
High-order cumulant tensors carry information about statistics of non-normally distributed multivariate data. In this work we present a new efficient algorithm for calculation of cumulants of arbitrary orders in a sliding window for data streams. We show that this algorithm offers substantial speedups of cumulant updates compared with the current solutions. The proposed algorithm can be used for processing on-line high-frequency multivariate data and can find applications, e.g., in on-line signal filtering and classification of data streams. To present an application of this algorithm, we propose an estimator of non-Gaussianity of a data stream based on the norms of high order cumulant tensors. We show how to detect the transition from Gaussian distributed data to non-Gaussian ones in a data stream. In order to achieve high implementation efficiency of operations on super-symmetric tensors, such as cumulant tensors, we employ a block structure to store and calculate only one hyper-pyramid part of such tensors.
2
Content available The system of streaming and analysis of signals from MEMS accelerometers
Fabiański B., Nowopolski K., Wicher B.
Poznan University of Technology Academic Journals. Electrical Engineering
|
2016
|
No. 87
301--312
EN
In the article, a dedicated testing environment for MEMS acceleration sensors is shown. The system is able to collect data from multiple devices with different physical interfaces, send them through parallel streaming, archive, and analyze it. The architecture and operational algorithms of individual components, such as complex synchronization methods in the data streaming process is described. This data streaming is finally realized by Ethernet interface which becomes a bridge between the PC system running the dedicated application and the sensor board. In the last section of the article, quality indicators of acceleration sensors signals are presented. These indicators indicate primarily a useful signal to noise ratio with respect to the measurement resolution.
3
Content available Selected algorithms of MEMS accelerometers signal processing in burglary detector application
Fabiański B., Nowopolski K., Wicher B.
Poznan University of Technology Academic Journals. Electrical Engineering
|
2016
|
No. 87
267--278
EN
In the paper, implementations and results of operation of artificial neural network applied as a burglary classifier are presented in comparison to solution with a direct digital signal processing (DSP) approach. The neural network operates in a mobile access control device, that may be easily attached to a door. The device is an integrated system, equipped with several sensors based on microelectromechanical systems (MEMS) technology. Due to limited effectiveness of simple, conditional logic algorithms on acquired signal samples, a more sophisticated approaches are investigated. Data acquisition during imitation of various burglary scenarios and further processing of the recorded signals are described in the paper. Selection of the neural network structure and pre-processing methods of sensor signals are presented as well. The direct DSP algorithm based on the application of the properties of application phenomena is shown in the same way. Finally, results of selected algorithms implementation in a low-power 32-bit microcontroller system are presented. Limitation of the platform responsiveness in the real-time conditions and comparison of used classification methods are discussed in the paper conclusions.
4
Content available The system of MEMS sensors data streaming and signal quality analysis
Fabiański B., Nowopolski K., Wicher B.
Computer Applications in Electrical Engineering
|
2016
|
Vol. 14
328--339
EN
In the article, a dedicated testing environment for MEMS sensors is presented. The system serve real–time measurements from several, different interfaced sensors, what gives opportunity to collect the data and – furthermore – its off–line analysis. To complete the main challenge what is MEMS ICs integration in one platform, a special hardware layer is applied together with operational algorithms. Two low–level boards are connected to the embedded server by RS–485 lines. This data server translates RS–485 signals and communicates with dedicated PC program by an Ethernet interface. Such a solution made possible to parallel streaming, archive, and analyze of data in a convenient way. The architecture and operational algorithms of individual components, such as complex synchronization methods in the data streaming process is described. Proper system design is verified by presenting selected signal waveforms grabbed in an experimental tests. In the end introduced two signal quality indicators resulting in comparison of different MEMS ICs. Summary table of computed indicators is shown with its analysis.
5
Content available Burglary detection based on accelometric data using selected signal processing algorithms
Fabiański B., Nowopolski K., Wicher B.
Computer Applications in Electrical Engineering
|
2016
|
Vol. 14
313--327
EN
The paper presents two approaches to the problem of burglary detection. The first one utilizes direct signal processing, while the other – artificial neural network (ANN). Both algorithms are compared in real operating conditions. The implementation of the algorithms was performed in a portable, battery operating devices that can be easily attached to the door. For direct comparison, two identical devices including several MEMS accelerometers and 32 bit microcontroller have been used – each with one algorithm implemented. The goal of using artificial neural network algorithm was to improve the performance of the burglary detection system in comparison to classical direct signal processing. The structure of ANN and required pre – processing of the input data, is presented and discussed as well. The article also describes the research system required to collecting the data for ANN training and to directly compare both algorithms. Finally, the results of behavior of the classification methods in real actual conditions is discussed.
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.