The lower limb exoskeleton is a wearable human–robot interactive equipment, which is tied to human legs and moves synchronously with the human gait. Gait tracking accuracy greatly affects the performance and safety of the lower limb exoskeletons. As the human–robot coupling systems are usually nonlinear and generate unpredictive errors, a conventional iterative controller is regarded as not suitable for safe implementation. Therefore, this study proposed an adaptive control mechanism based on the iterative learning model to track the single leg gait for lower limb exoskeleton control. To assess the performance of the proposed method, this study implemented the real lower limb gait trajectory that was acquired with an optical motion capturing system as the control inputs and assessment benchmark. Then the impact of the human–robot interaction torque on the tracking error was investigated. The results show that the interaction torque has an inevitable impact on the tracking error and the proposed adaptive iterative learning control (AILC) method can effectively reduce such error without sacrificing the iteration efficiency.
Lower limb exoskeletons (LLEs) are sets of mechanical devices used to support the action of human lower limbs. This recently developed technology has unprecedented potential in the construction industry by increasing the strength, endurance, and other physical capabilities of construction workers. For safety considerations, LLEs need reliable and responsive controllers to closely match their mechanical operation with human gait in a synchronous manner. This research proposes the use of physical human–robot interactive (pHRI) controllers that are suitable for construction tasks. The proposed pHRI integrates a gait trajectory‐based musculoskeletal model with iterative control algorithms. To minimize the trajectory tracking error between LLEs and human lower limbs, the gait dynamic was modeled as a spring damping and impedance model for supporting and swing phases. An iterative adaptive controller was developed for trajectory tracking and predication. To validate the proposed model, an in‐lab experiment to simulate typical construction activities was conducted, allowing us to assess tracking error. The experiment results suggest that the proposed model can minimize the trajectory tracking error to a level acceptable for safe operation. The iterative controllers allow fast error convergence for different construction scenarios with proper calibration. Therefore, the proposed pHRI iterative controllers are reliable and suitable for complicated activities within the dynamic working conditions intrinsic to construction sites.
Stability control is critical to the exoskeleton robot controller design. Considering the complex structural characteristics of lower limb exoskeleton robots, the major challenge of the controller design is the accuracy and uncertainty of the dynamics model. To fill in this research gap, this study proposes successive approximation-based radial basis function (RBF) neural networks (NNs). The proposed model simplifies the lower limb exoskeleton robot as three degrees-of-freedom (3-DOF) model with the human hip joints for adduction/extension, bending/extension, and internal/external rotation. To minimize the gait tracking errors and stabilize the closed-loop system, a gait trajectory-based control and approximation model was proposed in this study. To verify the proposed method, a validation experiment was conducted for typical lower limb motions. The experiment results demonstrated the effectiveness of the proposed method.
The lower limb exoskeleton is a wearable device for assisting medical rehabilitation. A classical lower limb exoskeleton structures cannot precisely match the kinematics of the wearer’s limbs and joints in movement, so a novel anthropomorphic lower limb exoskeleton based on series–parallel mechanism is proposed in this article. Then, the human lower limb movements are measured by an optical gait capture system. Comparing the simulation results of the series–parallel mechanism with the measured human data, the kinematics matching model at the hip joint is established. The results show that the kinematic matching errors in the X, Y, and Z directions are less than 2 mm. So, the proposed kinematics matching model is effective and the anthropomorphic series–parallel mechanism has a significant improvement in tracing the human positions at the hip joint.
This paper presents a hybrid adaptive approximation-based control (HAAC) strategy for a class of uncertain robotic joints' system. The proposed control structure consists of a robust sliding mode controller and an adaptive approximation-based controller. The robust sliding mode controller is designed by using the super-twisting algorithm, which is a particularly effective method to decrease the chattering caused by the traditional sliding mode control (SMC) and compensate the disturbances. Another improvement of the robust sliding mode controller is that the robust control parameters only subject to the upper bound of the derivative of the external disturbances, rather than choosing a relatively large value. Moreover, the designed adaptive approximation-based controller has the following two distinctive features: 1) the control parameters are designed to be adjusted in real time and 2) the prior knowledge of actual robotic model is not required to be known. These features contribute to compensating the uncertainties. The stability of the closed-loop system is proved by using the Lyapunov theory, and the simulation results demonstrate the effectiveness of the proposed control method. Finally, the proposed HAAC could apply in the experiments of industrial robotic joints' system. INDEX TERMS Hybrid adaptive control, robust sliding mode control, approximation-based control, robotic joints system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.