Explicit model based fuzzy control method for lower limb exoskeleton robot

Lower limb exoskeleton robots are vital tools for assisting in the rehabilitation and daily training of hemiplegic patients. However, achieving precise control of exoskeleton robots is a big challenge due to strongly nonlinear dynamics (e.g. gravity center variation, variation in gait, strap connection deviation), complex interaction between humans and robots, and disturbance. To address this issue, an explicit model based fuzzy control method is proposed for lower limb exoskeleton robots. In this method, an explicit model is initially derived based on Lagrange mean value theorem. This model does not involve any approximation for the dynamics of a nonlinear system and exhibits a linear structure, facilitating its use in designing controllers for such systems. Consequently, it effectively establishes the relationship between joint angles and control torques in complex environments. Then, a low-dimension fuzzy approach is developed for real-time parameter estimation of the explicit model. With this explicit model, a robust control law is derived. The performance analysis, stability proof, and experiments conducted on two types of exoskeleton robots-tested on both healthy individuals and patients-demonstrate that the proposed control method improves control accuracy by 47% and 22% compared to the traditional PID and SMC control methods, respectively.