A new dynamic model applied to electrically stimulated lower limbs and switched control design subject to actuator saturation and non-ideal conditions

• Autor(a) principal: Nunes, Willian Ricardo Bispo Murbak
• Data de Publicação: 2019
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://hdl.handle.net/11449/183168
• topic: Electrical stimulation; Paraplegics; Takagi-Sugeno fuzzy models; Linear matrix inequalities; Switched control; Estimulação elétrica; Paraplégicos; Modelos fuzzy Takagi-Sugeno; Desigualdades matriciais lineares; Controle chaveado

Electrical stimulation is a promising technique for motor rehabilitation in cases of spinal cord injury. Stimulator saturation is important in the control system designs applied to electrical stimulation. The negligence of the actuator saturation in the electrical stimulation can lead to unwanted control results, which evidences the muscular fatigue effects. For the first time a switched controller subject to actuator saturation for electrically stimulated lower limb is proposed. The dynamic limb extension model is nonlinear and uncertain. The uncertain nonlinear system described by Takagi-Sugeno fuzzy models operating within an operating region in the state space is considered in this study. In addition, fault in the actuator, muscle activation uncertainty, and muscular non-idealities, such as fatigue, spasms, and tremor were considered at three three severity levels. The switched controller is compared to parallel distributed compensation technique. Simulations denote better results of the switched controller by dealing with parametric uncertainties. On the other hand, a challenge for FES control systems is to monitor torque in muscle contractions. In isotonic contraction applications, measuring torque is difficult. The novelty in this study is the proposal of a new nonlinear model, whose state variables are angular position, angular velocity and angular acceleration. In this new model the torque variable is replaced by the angular acceleration. Experimental tests list the parameters corresponding to 24 individuals (20 healthy and 4 paraplegic) for the linearized model using gray box identification approach.