Academia.eduAcademia.edu

Figure 1 – uploaded by Abdulhamit Subasi

See full PDFdownloadDownload figure
Table 1 Class distribution of the samples in the training and test data sets

Table 1 Class distribution of the samples in the training and test data sets

Related Figures (4)

Fig. 1 a Fresh EMG signal, b STFT, ec SPWVD, d CWT of the signal For solving pattern classification problem MLPNN employing Levenberg—Marquardt (L-M) and gradient descent In this study, we propose the use of the Time—Frequency methods with ICA as an alternative method for the classification of the surface electromyography signal for
Fig. 1 a Fresh EMG signal, b STFT, ec SPWVD, d CWT of the signal For solving pattern classification problem MLPNN employing Levenberg—Marquardt (L-M) and gradient descent In this study, we propose the use of the Time—Frequency methods with ICA as an alternative method for the classification of the surface electromyography signal for
Fig. 2. a Fatigued EMG signal, b STFT, e SPWVD, d CWT of the signal The MLPNN was designed with different features of EMG signal in the input layer; and the output layer consisted of one node representing whether fatigue detected or not. A value of “0” was used when the experimental investigation indicated a fresh and “1” for fatigued muscle. Depending on the nature of the method, we used different nodes in the input layer for each method (STFT, SPWVD and CWT). We used 150 nodes for STFT, 160 for SPWVD and 175 for CWT. The preliminary architecture of the network was examined using one and two hidden layers with a variable number of hidden nodes in each. It was found that one hidden layer is adequate for the problem at hand. Thus the sought network will contain three layers of The adequate functioning of MLPNN depends on the sizes of the training set and test set. During training, the input and desired data will be repeatedly presented to the network. When using neural network, decision must be
Fig. 2. a Fatigued EMG signal, b STFT, e SPWVD, d CWT of the signal The MLPNN was designed with different features of EMG signal in the input layer; and the output layer consisted of one node representing whether fatigue detected or not. A value of “0” was used when the experimental investigation indicated a fresh and “1” for fatigued muscle. Depending on the nature of the method, we used different nodes in the input layer for each method (STFT, SPWVD and CWT). We used 150 nodes for STFT, 160 for SPWVD and 175 for CWT. The preliminary architecture of the network was examined using one and two hidden layers with a variable number of hidden nodes in each. It was found that one hidden layer is adequate for the problem at hand. Thus the sought network will contain three layers of The adequate functioning of MLPNN depends on the sizes of the training set and test set. During training, the input and desired data will be repeatedly presented to the network. When using neural network, decision must be
Table 2 Comparison of time-frequency methods for muscle fatigue detection using L-M algorithm
Table 2 Comparison of time-frequency methods for muscle fatigue detection using L-M algorithm
Table 3 Comparison of time-frequency methods for muscle fatigue detection using GDA algorithm
Table 3 Comparison of time-frequency methods for muscle fatigue detection using GDA algorithm
Related topics:
Information SystemsAlgorithmsSkeletal muscle biologyIndependent Component AnalysisTime-Frequency AnalysisPublic health systems and services researchMuscle Fatigue
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved