作者: M. Yuwono , A. M. A. Handojoseno , H. T. Nguyen
DOI: 10.1109/IEMBS.2011.6090760
关键词:
摘要: For people with severe spine injury, head movement recognition control has been proven to be one of the most convenient and intuitive ways a power wheelchair. While substantial research done in this area, challenge improve system reliability accuracy remains due diversity tendencies presence artifacts. We propose Neural-Network Configuration which we call Augmented Radial Basis Function (ARBF-NN). This network is constructed as (RBF-NN) Multilayer Perceptron (MLP) augmentation layer negate optimization limitation posed by linear classifiers conventional RBF-NN. The RBF centroid optimized through Regrouping Particle Swarm Optimization (RegPSO) seeded K-Means. trial results ARBF-NN on Head-movement show significant improvement up 98.1% sensitivity.