Robust Impedance Control for Lower-limb Rehabilitation Robot

Introduction: Nowadays, neurological impairments such as Stroke and Cerebral Palsy are the commonest forms of disability in adults and children respectively in the world. People with neurological impairment often have compromised volitional control of their arm and legs limiting their ability to undertake activities of daily living such as walking. The core mechanisms of rehabilitation interventions to promote lower limb function involve intensive practice of functional tasks, which drives neural plasticity to improve motor skills. However weakness of the lower limb makes it difficult for the neurologically impaired to practice at the necessary intensity. One-way of providing the support required for the neurologically impaired is to utilise rehabilitation robotic technology to enable practice of useful lower limb movements. This technology unlike commercially available systems enables the control of the spatial and temporal characteristics of movements. Methodology: One of the common control methods used is torque-based impedance control. This paper presents an Electromyogram-based robust impedance control for a lower-limb rehabilitation robot using a voltage strategy. The proposed control uses Electromyogram (EMG) signal in place of force sensor to estimate the force. In addition, the proposed control is based on the voltage control strategy, which differs from the common torque control strategy. To obtain patient forces, EMG signals are used to estimate real force, because we argue that force sensors can only measure contact force. Experimental EMG-force data is collected and used to train an artificial neural network. Furthermore, we design an adaptive fuzzy system to estimate and compensate the uncertainty in performing the impedance rule. The adaptive fuzzy system has an advantage that does not need new feedback to estimate the uncertainty. Results: To illustrate the effectiveness of the control approach, a 1-DOF lower-limb rehabilitation robot is designed. Simulation results show that compared with a torque-based control approach, the voltage-based is simpler, less computational and more efficient while it considers the presence of actuators. The control approach is verified by stability analysis. Simulation results show the efficiency of the control approach in performing some therapeutic exercises. Discussion: Impedance control is a very effective control for the rehabilitation robots. The previous impedance control approaches were developed based on the torque control strategy whereas the proposed impedance control is based on the voltage control strategy. The proposed approach is free from models for manipulator and patient, thus it is simpler, less computational and more robust and effective compared with the torque based control approaches. The fuzzy adaptive system has been efficiently used to adapt the controller to overcome uncertainties. The control approach has been verified by stability analysis. Simulation results show the superiority of the proposed adaptive impedance control approach over a torque based impedance control approach.