ANFIS and MPC controllers for a reconfigurable lower limb exoskeleton

The exoskeletons are robotic active orthoses intended to enhance power or in medical applications as rehabilitation and assistive walking. In the context of designing a controller for the reconfigurable exoskeleton for lower limb, it is critical to define the hardware as well as the control. The scope of this work was to define the controller for reconfigurable exoskeleton using three types of controllers: PD, ANFIS, and MPC. The PD controllers are the typical approach for torque/tracking control while artificial intelligence controller, as ANFIS, and optimal controller, as MPC, are recently entering to this field. The ANFIS and MPC controllers may bring more precision and capability to distribute the processing operations. Afterward, this work contrasts the performance evaluation using objective indices to evaluate the error, ISE, IAE, ITSE, ITAE, ISTSE, and ISTAE. The results suggest that ANFIS and MPC controllers have the potential to drive the torque/traction reducing the error while having the capability to learn from the disturbances from the surroundings.