Structural Analysis of Electrical Signals with Recurrent Use of a Multilayer Perceptron

Andreev, Oleg N.; Slavutskiy, Alexandr L.; Алексеев, В. В.

doi:10.3103/s1068371222080028

Cited by 4 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With additional signal processing and the perceptron recurrent use, the accuracy can be significantly increased [14]. As follows from the results above, the tracking currents and voltages signal parameters accuracy significantly depends both on the ANN structure and on its training mode.…”

Section: Discussionmentioning

confidence: 93%

See 1 more Smart Citation

Neural net monitoring of signals parameters during the induction motors run-out

Andreev,

Grigoriev,

Malinin

et al. 2024

Third International Conference on Optics, Computer Applications, and Materials Science (CMSD-III 2023)

View full text Add to dashboard Cite

The study is devoted to the possibility of using neural network algorithms to monitor the current and voltage signals parameters during transients in electrical engineering complexes. The direct propagation neural network can be programmed in standard microprocessor equipment and allows approximating signals in real time in the "sliding time window" mode. The proposed approach is demonstrated in laboratory experiments on the asynchronous motors run-out physical model as a complex load node part. It is shown that the multilayer perceptron use makes it possible to track changes in the signals frequency and amplitude with percentage accuracy for the observation time window duration of no more than a industrial frequency quarter period. The results demonstrate the using neural networks advantages for signal processing in non-stationary modes in electrical engineering and electric power industry.

show abstract

Section: Discussionmentioning

confidence: 93%

“…Many studies have been devoted to modeling such processes [9][10][11][12][13]. Such signal changes monitoring can be carried out in real time [14][15]. Artificial neural networks (ANN) can be used to solve this problem [16][17][18].…”

Section: Introductionmentioning

confidence: 99%

Neural net monitoring of signals parameters during the induction motors run-out

Andreev,

Grigoriev,

Malinin

et al. 2024

Third International Conference on Optics, Computer Applications, and Materials Science (CMSD-III 2023)

View full text Add to dashboard Cite

show abstract

“…First, using the ANN, for example, the gain factor k is determined, for which the learning error is much lower, then the time constant T , and only after that the damping coefficient  is calculated. The corresponding algorithm of perceptron using was tested in [26]. It is shown that the results can significantly increase the accuracy of determining the signal parameters.…”

Section: Resultsmentioning

confidence: 99%

“…The example described above addresses the main problem of the so-called "deep learning" of ANN. Deep does not exclude significant errors when using the ANN in the "if then" mode [1,26]. In this regard, the advantage of simple neural network algorithms is the ability to check their functioning at each stage of calculations [26,30].…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Neural net without deep learning: signal approximation by multilayer perceptron

Славутский

Lazareva²,

Portnov³

et al. 2023

2nd International Conference on Computer Applications for Management and Sustainable Development of Production and Industry (CM

View full text Add to dashboard Cite

The research is devoted to the use of artificial neural networks (ANN) for signal processing. The features of the simplest feed forward neural networks (multilayer perceptrons, MLP) application are analyzed. When using MLP in a sliding time window, it allows to solve problems of signal approximation with high accuracy and to determine their parameters when analyzing dynamic processes. If the signal can be set by analytical formulas with random parameters on separate time intervals, then after training, MLP can be implemented in microprocessor equipment for real-time signal processing. The ANN training algorithms and the errors of the proposed signal processing method are discussed. The approach does not require "deep learning" and a complex ANN structure, it allows one to control the accuracy of algorithms at intermediate stages of calculations. The results are of interest for electrical engineering and smart energy systems.

show abstract