Residual traveling distance estimation of an electric wheelchair

• Published in: 2012 5th International Conference on Biomedical Engineering and Informatics pp. 790 - 794
• Main Authors: Pei-Chung Chen, Yong-Fa Koh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2012
• Subjects: Fuzzy system; Neural networks; Residual traveling distance; Wheelchair
• ISBN: 9781467311830; 1467311839
• DOI: 10.1109/BMEI.2012.6513075

Based on economic considerations and estimation accuracy of wheelchair residual traveling distance, the virtual frictional force and virtual residual energy concepts are proposed in this paper. A virtual residual energy estimation system, based on fuzzy neural networks, is proposed using battery state of charge, wheelchair traveling speed and virtual frictional force as the inputs to estimate the wheelchair's virtual residual energy, and then transforms into wheelchair's residual traveling distance. A self-developed electric wheelchair using lithium battery as the energy source is employed to evaluate the proposed approach. The best estimated result, based on the root mean square error of estimated virtual residual energy, is 0.00573, while the worst one is 0.02182. On the other, the best estimated result, based on the root mean square error of residual traveling distance, is 0.402km, while the worst one is 1.285km. Thereby, the proposed estimation approach is feasible and can be applied to active vehicles.