Study on Lower Extremity Exoskeleton Signal Detection

The wave-filtering characteristics of the RLS self-adapting filter and the method of wavelet de-noising are studied, and the two are used as the methods of filtering noises in the fact signals tested from the lower extremity exoskeleton. In order to get better effects of filtering noises, different methods are used to different signals. The noise-filtering method based on changing the threshold of small wave transformation is used to filter white noise in digital signals. The self-adapting filter is used to other noises whose statistic characteristics are uncertain, or the both are combined to filter noises in such signal. For the wavelet de-noising of changeable threshold quality, there are two pivotal questions need to be settled. One is the choice of the threshold quality, the other is the determination of level-decomposition. In this paper, the adaptable determination method of level-decomposition is used, and the threshold quantity selection based on 3 sigma theorem is applied. The simulations of the fact signals tested from the lower limbs´ exoskeleton validated the superiority of noise-eliminating methods given in this paper.