A Multi-model LADRC Method for FES-based Upper Limb Rehabilitation

FES is an effective rehabilitation method for stroke patients, but joint angle variations influence muscle dynamics, leading to reduced rehabilitation performance. This paper proposes a multi-model LADRC-based control method for FES-based upper limb rehabilitation. Hammerstein musculoskeletal models are established for different angle ranges, with corresponding LADRC controllers designed. A multi-model switching mechanism selects the optimal controller based on real-time angle detection, while smoothing reduces abrupt signal variations. Simulation results demonstrate that compared to single-model LADRC, the proposed method significantly improves tracking accuracy and control stability, validating its effectiveness in FES-based upper limb rehabilitation.