ABSTRACT
The advancements in the field of cognitive neuroscience and human–machine interaction techniques have improved the capability of controlling devices with the human brain via a brain–computer interface (BCI) system. Specifically, the design of the brain-controlled mobile robots is essential to aid patients affected by neuromuscular illness. To achieve this, a new robot movement control technique is presented in this study using electroencephalography (EEG)-based BCI with an optimized kernel extreme learning machine (KELM), called BCI-OKELM. The presented model controls the movement of the robots based on the blinking of the operator’s eye through the EEG-based BCI, which makes use of alpha brain waveforms. The BCI-OKELM model comprises different stages of signal acquisition, preprocessing, feature extraction, and classification. Primarily, signal preprocessing is performed by removing the noise using filtering techniques and min-max normalization is used. In addition, the OKELM model is applied as a feature extractor followed by the inbuilt softmax (SM) layer for the classification process. For tuning the hyperparameters of the KELM model, an improved particle swarm optimization algorithm is employed. To investigate the betterment of the BCI-OKELM model, diverse simulation experiments are performed and the experimental results show the superior outcome with the sensitivity, specificity, precision, and accuracy of 81.36%, 87.24%, 76.85%, and 85.70% respectively.
