Neuromuscular electrical stimulation (NMES) has been effectively applied in many rehabilitation treatments of spinal cord injured (SCI) individuals. In this context, we introduce a novel robust and intelligent control-based methodology to closed-loop NMES systems. Our approach uses a control law to guarantee the system's stability. And, machine learning tools for both optimizing the controller parameters and system identification, with the novelty of using past rehabilitation data. In this paper, we apply the proposed methodology to the rehabilitation of lower limbs using a control technique namely robust integral of the sign of the error (RISE), an off-line improved genetic algorithm optimizer, and neural network models. Although in the literature the RISE controller presented good results on healthy subjects without any fine-tuning method, a trial and error approach would quickly lead to muscle fatigue for SCI individuals. Therefore, in this paper, for the first time, the RISE controller is evaluated with two paraplegic subjects in one stimulation session. And, with seven healthy individuals during at least one session up to at most five ones. As shown in results, control performance is improved via the proposed approach comparing to an empirical tuning, which can avoid premature fatigue on clinical procedures using NMES.

中文翻译：

---

一种通过神经肌肉电刺激进行人类下肢康复的新型鲁棒和智能控制方法

神经肌肉电刺激 (NMES) 已有效应用于脊髓损伤 (SCI) 个体的许多康复治疗。在这种情况下，我们为闭环 NMES 系统引入了一种新颖的基于鲁棒和智能控制的方法。我们的方法使用控制律来保证系统的稳定性。并且，用于优化控制器参数和系统识别的机器学习工具，具有使用过去康复数据的新颖性。在本文中，我们使用一种控制技术，即误差符号的鲁棒积分 (RISE)、离线改进遗传算法优化器和神经网络模型，将所提出的方法应用于下肢康复。尽管在文献中 RISE 控制器在没有任何微调方法的情况下在健康受试者上表现出良好的结果，反复试验的方法会很快导致 SCI 患者的肌肉疲劳。因此，在本文中，RISE 控制器第一次在一个刺激会话中对两个截瘫受试者进行评估。并且，在至少一个会话期间有七个健康个体，最多五个会话。如结果所示，与经验调整相比，通过所提出的方法提高了控制性能，这可以避免使用 NMES 的临床程序过早疲劳。