Predictor-Based Current Limitation Method for a DC Motor-Actuated Upper-Limb Rehabilitation Exoskeleton

• Published in: IEEE access Vol. 10; pp. 81684 - 81698
• Main Authors: Pont-Esteban, David, Sanchez-Uran, Miguel Angel, Contreras-Gonzalez, Aldo Francisco, Ferre, Manuel
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Brushless DC motors; Current limiters; D C motors; DC motors; Electric motors; embedded software; Exoskeletons; Hardware; Human engineering; Hysteresis motors; Overheating; Predictive models; Rehabilitation; rehabilitation robotics; Robots; Torque
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3194911

Current limitation is crucial to protect DC motors from overheating in overload situations. This consideration is more critical in applications that involve the movement of variable loads, such as exoskeleton applications, where there exists dynamic human-robot interaction. This work presents a software current limitation method based on a continuous DC motor model predictor. By predicting, according to the motor model, the voltages that would make the current match a threshold, the proposed method dynamically saturates the maximum voltages applicable to the motor to maintain the current under the predefined threshold. Using this method, a full-time current limitation can be obtained, but if desired, it also allows current peaks of configurable duration. Only motor parameters need to be known for the application of this algorithm, and it is easily adjustable to different DC motor models. Stand-alone motor tests show an effective limitation of current, achieving up to a 98.04% of maximum current when limited. The method has also been successfully validated in a rehabilitation exoskeleton application with four participants, obtaining an average current of 92.34% of current threshold when limited.