Fuzzy inference system for lower limbs angles prediction

This paper presents an approach of human gait modeling and prediction using fuzzy inference system. In this system, lower limbs motion prediction is simplified by introducing fuzzy knowledge base to describe the whole motion. Our fuzzy knowledge base predicts contralateral lower limb joints angles. The concept of fuzzy knowledge is based on fuzzy membership function that represents knee angle and differences between two consecutive angles and inference by IF-THEN rules. Selected fuzzy rules are emphasized by adding weights. Our final goal is to obtain predictive model of lower limbs movement in such way that it can generalize well with gait of different subjects. Prediction process is exemplified to validate the system in inter-subject generalization. The simulation results of newly designed system show the proposed method to be adequate and effective.