Electrically stimulated lower limb systems contain higher order nonlinearities and uncertainties in their physical parameters. Takagi-Sugeno (TS) fuzzy models are used to model nonlinear systems. Techniques such as parallel distributed compensation (PDC) are dependent on the membership functions that constitute the TS fuzzy model. When the exact representation approach is used to electrical stimulation applications, the system’s performance under PDC control can be deteriorated, because the membership functions may be uncertain, besides a high computational cost be required to compute them. In this paper, we propose a robust switched control subject to actuator saturation and fault (RSwASF) that effectively handles system uncertainties and nonidealities, such as fatigue, spasms, tremor, and muscle recruitment. Control techniques based on TS fuzzy modeling (PDC and robust PDC), as well as other approaches, such as sliding-mode control, backstepping, super-twisting, gain-scheduling, and proportional-integral-derivative (PID) control were compared to RSwASF through the root-mean-squared error (RMSE). The results indicate that RSwASF minimizes the influence of the parametric uncertainties and presents the lowest RMSE for healthy and paraplegic individuals.

电刺激下肢系统在其物理参数中包含高阶非线性和不确定性。Takagi-Sugeno (TS) 模糊模型用于对非线性系统进行建模。并行分布补偿 (PDC) 等技术依赖于构成 TS 模糊模型的隶属函数。当精确表示方法用于电刺激应用时，系统在 PDC 控制下的性能可能会恶化，因为隶属函数可能是不确定的，而且计算它们需要很高的计算成本。在本文中，我们提出了一种受致动器饱和和故障 (RSwASF) 影响的鲁棒切换控制，可有效处理系统不确定性和非理想情况，例如疲劳、痉挛、震颤和肌肉募集。将基于 TS 模糊建模（PDC 和鲁棒 PDC）的控制技术以及其他方法，例如滑模控制、反步、超扭曲、增益调度和比例积分微分 (PID) 控制与RSwASF 通过均方根误差 (RMSE)。结果表明 RSwASF 最大限度地减少了参数不确定性的影响，并为健康和截瘫个体提供了最低的 RMSE。