Reconstructing Bifurcation Diagrams of Induction Motor Drives Using an Extreme Learning Machine

Reconstructing one- and two-bifurcation diagrams of all components in the Rössler equations only from time-series data sets

“…In the BD reconstruction, time-series predictors are trained to model the obtained time-series data sets of all components, and the method allows the attractors of all components to be estimated when the bifurcation parameter values are changed. We reconstructed the BDs for various systems and numerical conditions using time-series data sets of a component in [8][9][10][11][12][13][14]. In this study, we show that the reconstruction requires a shorter length of training data when using time-series data sets of all components than when using the time series of only one component, compared by the previous studies.…”

“…Given its extremely high training speed and good generalization, we use the ELM as a time-series predictor. In previous work, we showed that the ELM is useful for reconstructing BDs [8][9][10][11][12][13].…”

“…Therefore, this method is suited to finding optimal parameters from several data sets. In a previous example, we demonstrated reconstruction of the BDs of induction motor drives [11]. To apply BD reconstruction to real-world data, we have shown that the BD reconstruction is robust against dynamical and observational noise because time-series data are always affected by noise [12,13].…”