China
Africa
Explore chapters and articles related to this topic
Utilisation of PID controller in explicit solver
Published in Alphose Zingoni, Insights and Innovations in Structural Engineering, Mechanics and Computation, 2016
J. Vorel, M. Marcon, R. Wendner, D. Pelessone, G. Cusatis
To smooth out the noise in the data obtained by the numerical simulation and used for the set up of the PID controller, a low pass filter is employed. There exist many different types of low pass filters which can be used to smooth the data. For the sake of simplicity and a low storage requirements, the exponential moving average (exponentially weighted moving average) is used in the proposed computational scheme. Moreover, to take into account possible inconsistent time steps, the exponential moving average for unevenly spaced time series is implemented (Eckner 2012, Müller 1991) () y¯(tk)=wy¯(tk−1)+(s−w)y(tk−1)+(1−s)y(tk),
Classifying unevenly spaced clinical time series data using forecast error approximation based bottom-up (FeAB) segmented time delay neural network
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2021
Y. Nancy Jane, H. Khanna Nehemiah, Arputharaj Kannan
The smoothing constant parameter for ‘α’ and ‘β‘ is updated dynamically based on the interval spacing between the observation. This dynamic adjustment overcomes the complexity of forecasting in unevenly spaced time series data. Table 3 shows the comparision of MSE, MAD, error rate and MAPE obtained for the classical DES, Wright updated DES and Hanzak updated DES . It was found that Hanzak updated DES shows high forecast accuracies compared to Wright and classical DES. The results indicate that the Hanzak updated DES shows high forecast accuracies compared to Wright and classical DES. Hence, Hanzak updated DES was used to compute forecasting accuracies for merge cost computation in FeAB segmentation. The segmented time series is then given to TDDN classifier to build a trained classification model. During the testing phase, the test set reserved using hold-out method is segmented using FeAB segmentation. The segments formed from the test set are used for testing the trained classification model.