Bending-Sensitive Optical Waveguide Sensor With Carbon-Fiber Layer for Monitoring Grip Strength.

With the gradual popularity of wearable devices, the demand for high-performance flexible wearable sensors is also increasing. Flexible sensors based on the optical principle have advantages e.g. anti-electromagnetic interference, antiperspirant, inherent electrical safety, and the potential for biocompatibility. In this study, an optical waveguide sensor integrating a carbon fiber layer, fully constraining stretching deformation, partly constraining pressing deformation, and allowing bending deformation, was proposed. The sensitivity of the proposed sensor is three times higher than that of the sensor without a carbon fiber layer, and good repeatability is maintained. We also attached the proposed sensor to the upper limb to monitor grip force, and the sensor signal showed a good correlation with grip force (the R-squared of the quadratic polynomial fitting was 0.9827) and showed a linear relationship when the grip force was greater than 10N (the R-squared of the linear fitting was 0.9523). The proposed sensor has the potential for applications in recognizing the intention of human movement to help the amputees control the prostheses.

Small-Data-Driven Temporal Convolutional Capsule Network for Locomotion Mode Recognition of Robotic Prostheses.

Locomotion mode recognition has been shown to substantially contribute to the precise control of robotic lower-limb prostheses under different walking conditions. In this study, we proposed a temporal convolutional capsule network (TCCN) which integrates the spatial-temporal-based, dilation-convolution-based, dyna- mic routing and vector-based features for recognizing locomotion mode recognition with small data rather than big-data-based neural networks for robotic prostheses. TCCN proposed in this study has four characteristics, which extracts the (1) spatial-temporal information in the data and then makes (2) dilated convolution to deal with small data, and uses (3) dynamic routing, which produces some similarities to the human brain to process the data as a (4) vector, which is different from other scalar-based networks, such as convolutional neural network (CNN). By comparison with a traditional machine learning, e.g., support vector machine(SVM) and big-data-driven neural networks, e.g., CNN, recurrent neural network(RNN), temporal convolutional network(TCN) and capsule network(CN). The accuracy of TCCN is 4.1% higher than CNN under 5-fold cross-validation of three-locomotion-mode and 5.2% higher under the 5-fold cross-validation of five-locomotion modes. The main confusion we found appears in the transition state. The results indicate that TCCN may handle small data balancing global and local information which is closer to the way how the human brain works, and the capsule layer allows for better processing vector information and retains not only magnitude information, but also direction information.

Anti-Disturbance Sliding Mode Control of a Novel Variable Stiffness Actuator for the Rehabilitation of Neurologically Disabled Patients.

Lower limb exoskeletons are widely used for rehabilitation training of patients suffering from neurological disorders. To improve the human-robot interaction performance, series elastic actuators (SEAs) with low output impedance have been developed. However, the adaptability and control performance are limited by the constant spring stiffness used in current SEAs. In this study, a novel load-adaptive variable stiffness actuator (LaVSA) is used to design an ankle exoskeleton. To overcome the problems of the LaVSA with a larger mechanical gap and more complex dynamic model, a sliding mode controller based on a disturbance observer is proposed. During the interaction process, due to the passive joints at the load side of the ankle exoskeleton, the dynamic parameters on the load side of the ankle exoskeleton will change continuously. To avoid this problem, the designed controller treats it and the model error as a disturbance and observes it with the disturbance observer (DOB) in real time. The first-order derivative of the disturbance set is treated as a bounded value. Subsequently, the parameter adaptive law is used to find the upper bound of the observation error and make corresponding compensation in the control law. On these bases, a sliding mode controller based on a disturbance observer is designed, and Lyapunov stability analysis is given. Finally, simulation and experimental verification are performed. The wearing experiment shows that the resistance torque suffered by humans under human-robot interaction is lower than 120 Nmm, which confirms that the controller can realize zero-impedance control of the designed ankle exoskeleton.